On the Pollen Harvest by the Honey Bee (Apis mellifera L.) Near Tucson, Arizona (1976-1981)

Item Type Article

Authors O'Neal, Robert J.; Waller, Gordon D.

Publisher University of Arizona (Tucson, AZ)

Journal Desert Plants

Rights Copyright © Arizona Board of Regents. The University of Arizona.

Download date 06/10/2021 16:00:38

Link to Item http://hdl.handle.net/10150/552253 O'Neal and Waller Honey Bee Pollen Harvest 81

Summary Near Tucson, Arizona, the pollen harvest by colonies of On The Pollen Harvest Honey Bee (Apis mellifera L.) was monitored with pollen traps. This influx was analyzed at 11 apiaries for 1 -3 years and by the Honey Bee (Apis 1 apiary for 5 years. Profiles of pollen influx show an apparent sensitivity to both the composition of nearby vegetation and mellifera L.) Near the seasonality of pollen production. The pollen influx from taxa most important in the pollen diet of colonies at one of the apiaries was diagrammed, providing the first detailed Tucson, Arizona account of pollen diets of Honey Bees. Most pollen could have been obtained within several (1976-1981) kilometers of the colonies, yet bees flew at least 12 km to obtain Brassica pollen. At another apiary, Manzanita (Arctostaphylos) pollen was collected 9 km from the hive. One colony on a mountain at 2438 m elevation utilized Robert J. O'Neall Mesquite (Prosopis), the nearest of which was 850 m lower in and Gordon D. Waller elevation and 5.6 km to the north. There were two major periods of pollen influx for colonies U.S.D.A., Agricultural Research Service foraging within the desert plant communities near Tucson: 1) Carl Hayden Bee Research Center February through June, and 2) August through October. Tucson, Arizonan Weights of pollen harvests made by colonies at one apiary for 4 years from 6 species of winter ephemerals were shown to be highly correlated with winter rainfall. The linear regression of total ephemeral pollen yield on total precipita- tion from November through April indicated that 76 mm of total precipitation had to be exceeded before pollen was harvested from the 6 ephemerals. Honey Bees utilized anemophilous pollen in this study to a far greater extent than would be expected for other bees. Stores of honey apparently enabled colonies to harvest such pollen since anemophilous plants are characteristically nectar- less. Anemophilous pollen contributed from 9% to 24% of the annual pollen harvest during three years at one apiary in this study. At another apiary, colonies depended upon anemophi- lous pollen for 71% of their pollen diet during the first half (April 1 to July 22) of the flowering season. The ability of the Honey Bee to exploit anemophilous pollen enlarged its poten- tial food base, insured greater reliability in pollen income, and probably increased the inhabitable range of this bee. Fluctuations of pollen harvests encountered in this study were compared with those at other latitudes. At high latitudes (45-65°) there proved to be a unimodal pulse in pollen availability during the year. At somewhat lower latitudes (30- 45°), the pulse was found to be bimodal. The data from temperate latitudes suggested that a global wave of potential flowering existed which completed one oscillation between poles each year, reaching the high latitudes once annually but passing through the mid -latitudes twice annually. Abundance- diversity curves of the annual pollen diet sets of Honey Bee colonies at Tanque Verde, and at 3 other localities in the northern temperate latitudes, each had log- normal distributions. The data sets from the Tanque Verde apiary, compared with published data from England, Wales, and Kansas, indicated overall greater diversity in the Tanque Verde pollen assemblages. In relative proportions contributed, 'Present address Smithsonian Tropical Research Institute, Apartado the two top- ranked species of each of the three other localities 2072, Balboa, Rep. de Panama, APO Miami 34002. contributed more to Honey Bee diets than did the two top - 2Mailing address: 2000 E. Allen Rd., Tucson, AZ 85719. ranked species at Tanque Verde. 82 Desert Plants 6 (2) 1984

Introduction Table 1. Most of these studies present either the seasonal Bees (Hymenoptera, Superfamily Apoidea), with the excep- influx of pollen or the proportion of pollen contributed by tion of some cleptoparasitic or necrophagous species (Schwarz, various plant species at each time interval. Only Synge 1948: 106, Roubik, 1982a), derive their nutrition from nectar, (1947), Percival (1947), and Rashad and Parker (1956) col- pollen and plant oils (Spencer- Booth, 1960; Wahl, 1963; lected data continuously for at least one year and, with the Haydak, 1970; Vogel, 1974). For most bees, pollen is the major pollen analyzed for each sample, presented estimates of the source of proteins, lipids, carbohydrates, vitamins and min- annual contribution by each plant species to the colonies' erals. Many species of bees, in order to provision their nests, pollen diet. harvest pollen from only one to several species of plants Both the total influx of pollen and the annual contribu- while the pollen diets of other bees, especially long -lived or tion by each plant taxon are useful bits of information for perennial highly social species, include many plant taxa comparative analysis. Yet, few comparative analyses have (Michener, 1979; Roubik and Michener, 1980). The bee with been made of the pollen harvests by Honey Bee colonies in the largest pollen diet breadth is the Honey Bee, Apis different years at the same locality, and no regionalcom- mellifera L.It harvests pollen from virtually all taxa of parisons have been made of Honey Bee pollen diets at spermatophytes (Schmalzel, 1980). With its present cosmo- different latitudes or in different types of vegetation. politan distribution it has access to most of the world's In this paper we examine the contribution that analysis terrestrial flora. of Honey Bee pollen diets (melittopalynology) can make to Not only does the Honey Bee utilize the largest spectrum the study of Honey Bee ecology and to an understanding of of plants for food of any species of bee, but it also harvests several parameters of flowering events. The results of large quantities of pollen and is one of the most ubiquitous monitoring closely the pollen diets of colonies in southern pollen feeders in the world today. Estimates of the amount of pollen consumed annually by a Honey Bee colony range from 15 to 55 kg (Spencer- Booth, 1960). In the United States alone, Table 1. Studies that have utilized pollen traps to monitor the where the population of Honey Bee colonies is at least 4.16 x pollen harvest by honey bees. 106 (U.S.D.A., 1.976), Honey Bees are harvesting between 62 Authors Location Type of Study' and 228 x 106 kg of pollen annually. Todd and Bishop (1940) California a Throughout most of the foraging season, the pollen brought into a Honey Bee colony is rapidly used in brood -rearing. This Eckert (1 942) California a is especially apparent from data on the Honey Bee in the Whitcomb (1944) Louisiana a tropics (Smith, 1960; Winston, 1980), the subtropics (Kauffeld, Stapel and Eriksen (1944) Denmark b 1980), and to a lesser extent in the temperate regions (Jebsen, 1952; Jeffree and Allen, 1957). If most pollen is rapidly Todd and Bishop (1946) California a converted to bee biomass within a colony, the rate of brood Hare and Vansell (1946) Utah a, b2 production should be positively correlated with the rate at Synge (1947) England c which pollen is harvested by the colony (pollen influx). Both Percival (1947) Wales c Nolan (1925: 25) and Betts (1928) indicated that pollen influx was largely responsible for observed brood cycles of Honey Anonymous (1947) England b Bee colonies. Harris and Filmer (1948) New Zealand b

Neoptropical social bees show similar patterns (Roubik, Vansell and Todd (1949) western U.S.A. a3 1982b). Todd and Bishop (1946) graphically showed that brood production followed the pollen influx at Davis, Hirschfeldcr (1951) Germany a California. Maurizio (1953) Switzerland c A pollen trap on a hive is the most direct means of Rashad and Parker (1956) Kansas c monitoring the influx of pollen into a Honey Bee colony. Louveaux (1958, 1959) France a The trap operates simply by dislodging pollen pellets (_ corbicular loads or bee baskets) from the hind legs of Moriya (1960) Japan b Honey Bees as they enter the hive. A variety of pollen traps Thompson (1960) Arkansas a have been invented (see review by Crane, 1976). Traps that Olsen (1975) Michigan b are highly efficient must be rotated about once a week Adams et al. (1978) Ontario b between colonies so that brood production in each colony can be sustained. A trap that intercepts about 50% or less Kauffeld (1980) Louisiana a of the incoming pellets does not appear to lower the Ramirez (1980) Costa Rica a colony's population (Waller et al., 1981) and can be used to Severson and Parry (1981) Wisconsin b continuously monitor the pollen influx into one colony for at least a year. The invention of the pollen trap occurred ' a: the seasonal influx of pollen into the pollen trap b: an analysis of the proportion of pollen contributed by various almost 50 years ago. It is curious that few records of pollen plant species at each time interval influx into Honey Bee colonies have been published, c: both a and b considering the obvious utility of influx records in apicul- 2 samples taken for only 22 days ture and in research on the foraging behavior and ecology of this animal. Studies that have used pollen traps to 3 most records probably incomplete monitor the pollen harvest by Honey Bees are listed in 4 pollen influx not methodically sampled each day 111 e

lo o 10 11111111111 210 Mi les 10 0 101 2.0 Illllllll u Kilometers

Figure 1. Locations of the apiaries in this study: RT = Ragged Top; CR = Cortaro Road; SP = Steampump; TV = Tanque Verde; SNM = Saguaro National Monument; SX = San Xavier Mission; ASDM = Arizona- Sonora Desert Museum; SH = Summerhaven; FV = Falcon Valley; and RJ = Robles junction. Base map compiled by U.S. Geological Survey. Contour interval 2500 ft. (762 m).

Arizona are presented. Then, the significance of anemophi- of the foraging radius of a colony (von Frisch, 1967), the lous pollen in the diet of the Honey Bee is discussed. And Honey Bees (considered collectively) in this study had finally, using available records of the seasonal pollen potential access to vegetation from 549 to 2790 m elevation. harvest by Honey Bee colonies, global patterns of floral This range of elevation encompasses vegetation as disparate availability are described. as the Sonoran Desert scrub that grows on the hot, dry valley floors and the fir forest that grows on the cool, moist Location and Description of the Study Areas mountain slopes. Honey Bee colonies were kept at 12 locations (apiaries), The following brief descriptions of 8 plant associations all of them within 60 km of Tucson, Pima County, Arizona available to the Honey Bees studied are adapted from Lowe (32°15'N, 111 °0'W). The locations of the apiaries are shown 1964). Shreve (1951), Lowe( 1964) and Whittaker and in Figure 1. The elevations of the apiaries range from 700 m Niering )1965) and references therein present a more de- to 2438 m (Table 2). If 12 km is regarded as the upper limit tailed description of the vegetation than can be introduced 84 Desert Plants 6 (2) 1984 here. The nomenclature of plants mentioned in this paper Wheat (Triticum aestivum), Cotton (Gossypium spp), Lettuce follows that of Lehr (1978), Kearney and Peebles (1960), (Lactuca sativa), and Alfalfa (Medicago sativa). Cultivation Munz (1959), Bailey ( 1973), and Bailey ¡1976). and flood irrigation are practiced throughout the year. The Creosotebush. There are presently three communities fields and irrigation ditches frequently support a luxuriant that characterize the Sonoran Desert in southeastern growth of disturbance plants including'Brassica ssp., London Arizona: Creosotebush, Mesquite -Burroweed, and Paloverde. Rocket (Sisymbrium irlo), Sow Thistle (Sonchus oleraceus), The Creosotebush community is a simple but extensive Wild Lettuce (Lactuca serriola), Barley (Tlordeum ssp.), and community confined to the broad, gently sloping valleys where well -drained, silty soils occur. Two shrubs, Creosote - bush (Larrea divaricata) and Triangle Bursage (Ambrosia Table 2. Elevations of the apiaries and the plant associations available to the bees. ( ) indicates an association with limited deltoidea), are the dominant perennial plants. During win- availability to the bees at a given apiary. ters with sufficient rainfall, ephemeral herbs such as Alfilaria (Erodium cicutarium), Scorpionweed (Phacelía spp), Range of Mexican Poppy (Eschscholtzia mexicana), Yellow Rocket Elevations Plant Associations ( Sisymbrium irio), Fiddleneck (Amsinckia intermedia), Elevation within 10km Available within Twist Flower (Streptanthus arizonicus), Desert Lupine Apiary (m) of Apiary 10 km of Apiary (Lupinus sparsiflorus) and Gordon Bladderpod (Lesquerella gordoni) are a conspicuous element of the Sonoran Desert, Ragged Top 700 550 -1290 Paloverde-Saguaro- El Paso Gasline Ironwood, Creosote - covering extensive areas of the desert. Harris Hawk bush, (Turpentine Mesquite -Burroweed. Mesquite (Prosopis velutina) Painful Bush) and Burroweed (Haplopappus tenuisectus) occur as an Cortaro Road 730 640 -1220 Paloverde- Saguaro -lron- association throughout much of the uplands (762 -1070 m wood, Irrigated elev.) of southeastern Arizona. Persistent grazing and soil Farmland, Creosote - trampling by cattle are recognized as having contributed to bush, (Riparian the depauperate nature of this association. Burroweed, Woodland) Snakeweed (Gutierrezia spp), Prickly Poppy (Argemone Steampump Ranch 760 670 -2410 Paloverde- Saguaro, spp.), and Bull- Nettle (Solarium elaeagnifolium) -all un- Riparian Woodland, palatable to cattle -grow among the scrubby Mesquite. Turpentine Bush, Chaparral, Oak -Pine The winter ephemerals found in the Creosotebush com- Woodland, Irrigated munity are also present in this community. Herbaceous Farmland plants found in this community that grow after the summer San Xavier Mission 762 730 -1090 Irrigated Farmland, rains include Tolguache (Daturaspp.),Devil's Claw Paloverde- Saguaro, (Proboscides spp), Spiderling (Boerhaavia spp), other species Mesquite -Burroweed, of Nyctaginaceae, Summer Poppy (Kallstroemia spp), Spurge (Riparian Woodland) Euphorbia spp.), Pig Weed (Amaranthus spp.), and Arizona -Sonora 792 670 -1220 Paloverde- Saguaro- Tidestroemia Ianuginosa. Desert Museum Ironwood, Creosote - Paloverde. On the rocky slopes along the base of the bush, (Irrigated Farmland, Turpentine mountains, a complex community of shrubs and dwarf Bush) shrubs dominates, with Foothill Paloverde (Cercidium microphyllum) and Saguaro (Carnegiea gigantea) as the Tanque Verde 790 760 -1830 Riparian Woodland, Creosote -bush, most conspicuous elements of the association. Ironwood Paloverde- Saguaro, (Olneya tesota), almost equally as abundant as Foothill Mesquite -Burroweed, Paloverde in southwestern Arizona, is well represented (Chaparral, Oak -Pine only near the Arizona -Sonora Desert Museum and the Woodland) Ragged Top apiaries, and to a lesser extent at the Robles Robles Junction 820 760- 920 Paloverde- Saguaro- Junction apiary. Saguaro and Ironwood are both absent Ironwood, Mesquite - from Paloverde communities at higher elevations, owing to Burroweed, (Irrigated Farmland, Turpentine their intolerance of prolonged freezing temperatures. Mes- Bush) quite (Prosopis velutina), Catclaw (Acacia greggii), White - Thorn (A. constricta), Blue Paloverde ( Cercidium floridum), Saguaro National 945 820 -1830 Paloverde- Saguaro, Monument East Creosotebush, and Desert Hopbush (Dodonaea viscosa) grow along the Mesquite -Burroweed, washes in the Paloverde community. Other shrubs in the Chaparral, Oak -Pine area include Jojoba (Simmondsia chinensis), Rhyolite Bush Woodland, (Riparian (Crossosoma bigelovii), Sangre de Cristo (Jatropha cardio- Woodland) phylla), Janusia gracilis, Paperflower (Psilostrophe cooperi), Falcon Valley 1127 1000 -1670 Paloverde, Mesquïte- Brittlebush (Encelia farinosa), Desert Zinnia (Zinnia Burroweed, Creosote - acerosa), Ocotillo (Fouquieria splendens) and Opuntia spp. bush, (Chaparral, Oak Woodland) Winter and summer ephemerals common to the other two Sonoran Desert communities are also found here. Summerhaven 2438 1060 -2790 Ponderosa Pine - Douglas Fir, Oak -Pine Irrigated Farmland. Large areas near the San Xavier Woodland, Chaparral, Mission apiary and the Cortaro Road apiary are used to grow Riparian Woodland O'Neal and Waller Honey Bee Pollen Harvest 85

Brome Grass (Bromus spp) during the spring and Sunflower Arizona Rosewood (Vauquelinia californica). Succulent and (Helíanthus spp.), Crown -Beard (Verbesina encelioides), semi -succulent liliaceous plants in the woodlands include Johnson Grass (Sorghum halepensis), Pig Weed (Amaranthus Agave schottii, A. palmed, Sotol (Dasylirion wheeleri), Yucca palmen ), Milk Thistle (Silybum marianum), Russian Thistle spp., and Beargrass (Nolina microcarpa). (Salsola kali), Bermuda Grass (Cynodon dactylon), Ground Ponderosa Pine- Douglas Fir. Above 1800 m eleva- Cherry (Physalis spp), Tolguache (Datura inoxia), and Bind- tion, the forests in the Santa Catalina Mountains are domi- weed (Convolvulus spp) during the summer. nated by Ponderosa Pine (Pinus ponderosa) and Douglas Fir Riparian Woodland. Best developed near Tanque Verde (Pseudotsuga menziesii). The forests above about 2275 m, in and Falcon Valley apiaries, the riparian woodland community addition to these species, contain White Fir (Alines concolor), is found along streambeds that frequently run and have a Aspen ( Populus tremuloides), and Southwestern White Pine water table close enough to the surface to be available to (Pinus reflexa). Small understory trees include Gambel Oak deep- rooted trees. The water table at Tanque Verde is only (Quercus gambelii), Net -Leaf Oak (Q. reticolata), Silver Leaf about 5 meters below ground level. Trees characteristic of the Oak (Q. hypoleucoides), Arizona Madrone (Arbutus arizonica), riparian woodland include Fremont Cottonwood (Populus and New Mexican Locust (Robinia neomexicana). Along the fremontii), Velvet Ash (Fraxinus pennsylvanica var. velutina), higher -elevation streams, where the canopy of conifers is Arizona Walnut (Juglans major), Western Soapherry (Sapindus open, grow Arizona Alder (Alnusoblongifolia), Willow (Salix saponaria var. drummondii), Canyon Hackberry (Celtis reticu- spp), Red Osier Dogwood (Cornus stolonifera), Cow Parsnip lata), Arizona Sycamore (Platanus wrightii), Willow (Salix (Heracleum lanatum), Cliff Bush (Jamesia americana), Mock spp), and arborescent Mesquite (Prosopis velutina). In and Orange (Philadelphus microphyllus), and New Mexican Rasp- along the streambeds during the summer grow Sunflower berry (Rubus neomexicanus). The roadsides and meadows (Helianthus annuus), Golden Eye (Viguiera spp), Clammy-weed have Sweet Clover (Melilotus spp), Clover (Trifolium spp), (Polanisia trachysperma), Jackass Clover (Wislizenia refracta), Vetch (Vicia spp), Mullein (Verbascum thapsus), Dandelion Scarlet Morning Glory (I pomaea coccinea), Desert Cotton (Taraxacum spp), Penstemon barbatus, Thistle (Cirsium spp), (Gossypium thurberi), Wild Bean (Phaseolus spp), and Cockle- Sneeze Weed (Helenium hoopesii), Macromeria viridiflora, bur (Xanthium saccharatum). The floodplains support shrubs Cilia sp., Crane's Bill (Geranium spp.), Lupine (Lupinus spp), such as Desert Willow (Chilopsislinearis), Seepwillow and Bee Balm (Monarda austromontana). (Baccharis glutinosa), Desert Broom (Baccharis sarothroides), and Burro Brush (Hymenoclea monogyra). The riparian wood- Methods and Materials land probably offers more nectar and pollen per hectare to Collection of Pollen Influx Records. At least 3 Honey Bees than any other community available to the bees Honey Bee colonies with pollen traps were kept at each of the in this study. Prior to the destruction of the Mesquite apiaries. The colonies used in this study usually had active woodland (bosque) along the Santa Cruz and Gila rivers, it queens, were free from disease, and had adult and brood popu- was common for beekeepers to keep 200 colonies at a single lations that were equivalent to colonies of normal size at the apiary. Thirty colonies is considered to be the upper produc- same apiary. Adult populations were rarely less than 15,000 tive capacity for an apiary in this region today. bees per colony. Colonies were kept in standard Langstroth Turpentine Bush. Turpentine Bush (Haplopappus lari- hives. Crevices in the hive bodies were sealed so that the bees ci folius) growing with Wild Buckwheats (Eriogonum spp., esp. could only use the entrance to the hive. At all times a min- E. wrightii and E. fasciculatum) can only be loosely regarded imum of 15 kg of honey was in each hive to serve as a source as a plant association separate from that of the oak woodland, of food for the colony. Colonies at all but one of the apiaries where they are both common. Yet because they are important made a surplus of honey; only the colonies at Summerhaven floral resources to Honey Bees and live independently of had to be supplied with honey. Pollen was not fed to the most oak woodland elements on the upper bajadas near colonies since previous studies (Free and Williams, 1971; several apiaries, they are considered separately. Both of these Moeller, 1972) had indicated that colonies fed pollen col- plants, in addition to Arizona Rosewood ( Vauquelinia califor- lected less pollen than colonies not fed. nica), Oreganillo (Aloysia wrightii), Milkweed (Asclepias Although several designs of pollen traps were available, linaria), and Sage (Salvia sp), survive in the foothills near only the modified Ontario Agricultural College (O. A. C.) Robles Junction apiary and on the north- facing slope of pollen trap (Smith and Adie, 1963; Waller, 1980) was employed Ragged Top, in the latter case, well outside the present range to recover a portion of the pellets of pollen that the bees of oaks. brought into the hive. The traps were kept on the colonies Chaparral and Oak -Pine Woodland. Only three api- continuously through the year. Preliminary data suggested aries (Summerhaven, Saguaro National Monument East, and that about 60% of the incoming corbicular pollen was Steampump Ranch) are in close proximity to these associa- removed by this trap (O'Neal, unpublished). This was inef- tions. Chaparral occurs between 1200 and 1800 m elevation. ficient enough that pollen entered the colony at a rate per- In the Santa Catalina and Rincon Mountains, chaparral is mitting colony development and maintenance only slightly predominantly composed of Manzanita (Arctostaphylos pun - less than that of untrapped colonies. gens and A. pringlei), Pachaba (Brickellia californica), and Silk Ants, if allowed to forage in the pollen traps, could have Tassel (Garrya wrightii). The oak -pine woodland intergrades removed significant amounts of pollen from May to with chaparral in these mountains and includes Mexican September, the months when they were most active. After Pinyon (Pinus cembroides), Bellota or Emory Oak (Quercus 1976, those hives having pollen traps were kept on stands emoryi), Arizona White Oak (Q. arizonica), Mexican Blue Oak that had legs in pans of oil to keep ants out of the pollen (Q. oblongifolia), Alligator Juniper ( Juniperus deppeana), and drawers. 86 Desert Plants 6 (2) 1984

Although it was desirable to empty the pollen traps at generic levels quickly since many grains that appear to be weekly intervals, this was not always accomplished. Most from the same species can easily be inspected and measured. collections were taken in intervals of less than 10 days. Another advantage to inspecting pellets rather than mixtures Because collections were usually made in the evenings, the of pollen grains is that pellets, should they be composed of intervals represented whole days. The collections were cleaned rare substances such as fungal spores or gymnosperm pollen, of insects and debris by hand, Additional cleaning of fine will not be considered incidental contamination and can be debris was done by sifting the pollen with a No. 18 U.S. examined more closely. Standard sieve (1.00 mm openings), The pollen was weighed The second assumption is that the probability of capture fresh. However, if the pollen was very moist during the infre- for a pollen pellet entering the pollen trap is constant. This quent periods of rainfall and high humidity, it was air dried in assumption is tenuous since the pellets are not equal in size the sun for several hours before cleaning. After 1978, the and the efficiency of the pollen trap is related to the size of collection in each trap was weighed separately. Previous to the pellet entering the trap. Large pellets have a greater this time the collections from one apiary were combined probability of being captured than small pellets. The analysis before weighing. is performed to determine the proportion of each pollen After weighing the pollen collections, the pollen from a taxon in the sample. Pellets of plants such as Eucalyptus and given apiary representing a given time interval was carefully Salt Cedar (Tamarix) are typically small (ca. 4 mg dry weight). mixed and a random sample of pollen pellets (about 50 g) was Saguaro (Carnegiea) and Dandelion (Taraxacum) pellets are drawn and stored frozen in a labeled plastic bag until ana- typically large (ca. 10 mg). Relative to the large-pellet taxa, lyzed. If pellets unusual in color or texture were noticed in taxa with smaller pellets will be under-represented in the the pollen collection, these were removed after sampling and pollen trap collection and over-represented in the analysis. stored separately. The degree of error introduced by the second assumption in The pollen influx into the colonies at a given apiary was the analysis is unknown, computed by calculating r, the mean rate of pollen harvested The determination of plant composition in a sample was by the pollen traps on the colonies: based on the microscopic examination of the pollen in each of 200 pellets randomly drawn from the storage sample. The pellets were initially color-sorted to decrease the total time for analysis by reducing manual activities of sorting and total amount of pollen collected recording individual pellets. Each pellet was individually (no. of colonies) (no. of days) inspected by chipping some pollen from the pellet and mixing the pollen into a drop of glycerin on a microscope slide. The pollen was viewed at 400X magnification; unfamil- iar pollen was viewed at 1000X using oil emersion. For more Thus, although in this paper pollen influx is noted as X detailed inspection of pollen grains or for photographic grams.colony-l.day-', the influx is based on the pollen trap documentation, pollen was acetolyzed using a modified harvest. All influx data given in this paper represent the method of Erdtman (1952). Pollen identification was facilita- grams of pollen trapped; adjustments were not made to esti- ted by developing a reference collection of the pollen of the mate the total amount actually collected by the colony. plants growing within several kilometers of the apiary and by referring to pollen floras by Kapp (1969), Heusser (1971), Analysis of Pollen Samples. Analytical procedures Huang (1972, Adams and Morton (1972, 1974, 1976, 1979), that were used to identify pollen to plant source and to McAndrews et al. (1973) and Markgraf and D'Antoni (1978). determine the relative importance of the different plants It is preferable when expressing the influx of pollen for the in the annual pollen diets of the colony are presented below. ith taxon to show its contribution to the total influx of pollen Two assumptions were made in the use of these procedures. (Ni), where pi is the proportion of the total number of pellets First, it was assumed that each pellet was a homogeneous in the sample consisting of the ith species, rather than mass of pollen from a single plant species. Maurizio (1953) expressing pi only in relation to 1-pi as previously done by found the proportion of mixed pellets (pellets consisting of Maurizio (1953), Severson and Parry (1981), and others. more than one pollen type) in a pollen collection to be typi- The importance of each plant taxon to the bees at the cally 1-3% with a range of 0.1-11.3%, This observation is in Tanque Verde apiary was determined by calculating Pi, the accord with Percival (1947) and with our own observations. proportion of taxon i in the annual diet of the colony: M.S. Brower (personal communication), however, has encoun- tered samples with as much as 60-70% of the pellets mixed. q However, we feel that our assumption of pellet homogeneity rjtipji j= is generally valid. This is convenient because the pellet, Pi rather than the pollen grain in the procedures of Chauvin and q Louveaux (1956) and Adams et al. (1978) can then be used as riti the unit of measure on which percentages are based. Pollen Hl grains are more readily identified from the homogeneous mass of a single pellet than individually from a pollen slurry where q is the total number of collections made in one year, r made from the total sample. This is because equatorial and is the rate of pollen influx (g-rams.colony4.day1) determined polar views of grains of the same pollen type can be viewed for the ¡th collection, ti is the number of days during which simultaneously. There is also opportunity to identify pollen the ¡th collection was made, and pis the proportion types belonging to taxa such as Pinus or Quercus to sub- contributed to the collection during tj by the ith taxon. O'Neal and Waller Honey Bee Pollen Harvest 87

Diversity of a sample, a function of both the number of maximum rates of pollen influx for the year when their species present in the sample and of the evenness with populations were at their lowest levels. At Summerhaven, the which individuals (pellets) are distributed among these spe- colonies were without available sources of pollen during the cies, was calculated using the Shannon -Weaver index of 5 months of winter. Yet on May 1st, one month after the diversity, H' (Shannon and Weaver, 1949). pollen harvesting resumed, the colony was collecting pollen at a rate of 0.5 kg/day and had already collected 7 kg of pollen Results and Discussion during April. Colonies were pollen -starved at the onset of the Flight Ranges. Considering the results of several kinds season and the profiles may have exaggerated theinitial of observations, von Frisch ( 1967: 66, 67( indicated that Apis availability of pollen. However, the reduced colony popula- mellifera obtained nectar from vegetation usually within 4 tions did not preclude pollen harvests of the magnitude km of the colony and that 12 km would be an extreme experienced later in the season. distance for a bee to forage from its colony. There have been Pollen samples that were collected from 5 apiaries in 1976 no studies that specifically examine the maximum range during periods of pollen influx greater than 40 g.colony'.day I that pollen can be obtained by Honey Bees. This study is no were analyzed to determine which plants contributed tothe exception. However, several remarkable observations were major pulses of pollen influx (Table 4). The pollen samples made concerning the distances Honey Bees flew for pollen. collected from the other 7 apiaries during 1980 -1981 were One colony at the Painful apiary had only Brassica pollen only spot checked, except for Summerhaven samples taken in its trap during November, 1980. At the same time, the from April 1 to June 22, 1981, which were analyzed completely other colonies at Painful and at the other two apiaries at (Table 5). The analyses were sufficient to describe the most Ragged Top were harvesting pollen only from Euphorbia that general patterns of pollen influx. grew nearby. The nearest Brassica was B. luncea inirrigated In agreement with Todd and Bishop (1940), Synge (1947), fields and ditches at least 12 km from the colony. Manzanita and Percival (1947, 1955), the times of maximum pollen col- (Arctostaphylos) pollen was present in traps at Tanque Verde lection were generally coincident with the flowering of the apiary at the end of February, 1976. The nearest Arcto- most numerically abundant species near the apiary. Al- staphylos was in chaparral 9 km away. One mixed pollen though the rate of pollen harvested from a given species by pellet, obtained at the Summerhaven apiary in July of 1980, the colony generally seemed to match the rate of flower was composed of Prosopis, Mimosa and Lepidium.The bee production by that plant, field censuses of flower popula- had to fly 5.6 km north and descend 850 m to reach the tions were not undertaken to check this impression. Such a nearest patches of these plants. In these cases, Brassica, check would have been valuable, especially if the census Arctostaphylos, Prosopis, Mimosa and Lepidium were all had included estimates of total pollen production per hec- potentially good sources of nectar, so that a bee foraging on tare by the various species, in determining how well the these plants at a great distance from the colony could have bees tracked flowering events and what biases existed when replenished the calories it had expended in flight. Two cases suggest that the maximum distance for bees to obtain pollen from nectarless plants may be less than half the Table 3. The total amount of pollen collected per colony per year at distance that prevails when pollen is associated with nectar 11 apiaries near Tucson, Arizona. Annual yields for the Arizona - or when nectar alone is present. Populus was a majorpollen Sonora Desert Museum and the Robles Junction apiaries include source every year at Tanque Verde apiary, yet it was not interpolated yields for mid -July thru September. harvested by the bees at Saguaro National Monument apiary, Year Amount Harvested (g) even though extensive stands were available 6.5 km tothe Apiary north. The bees at Tanque Verde did not collect pollen from El Paso Gasline 1980 -81 9171 Jojoba (Simmondsia) which was abundant 6.5 km to the Harris Hawk 1980 -81 11540 northeast and southeast. Painful 1980 -81 17240 Seasonality and Sources of the Pollen Harvested Steampump Ranch 1976 12962 at the 12 Apiaries. Profiles of the pollen influx at the 12 San Xavier Mission 1976 14274 apiaries are presented in Figures 2 and 3 and the annual yields Arizona -Sonora 1980 -81 20698 in Table 3. The profiles for the apiaries at Ragged Top (Fig. 3; f, Desert Museum g, h) are quite similar in pattern. The similarity is aresult of 12713 the bees foraging on the same vegetation types since the Tanque Verde 1976 apiaries are within 3 km of each other. Each of the profiles of 1977 9045 the other 9 apiaries have characteristics suggesting that the 1978 15771 timing and extent of individual peaks are largely due to the 1979 19924 flowering activity and composition of vegetation surrounding each apiary. The profiles of the 10 apiaries below 1000 m 1980 19167 elevation indicate that a pollen harvesting (= flowering) 5-year X ± 1 SD 15324 X 4538 season of 9.5 months is available to these colonies. At Falcon Robles Junction 1980-81 16090 Valley apiary (1127 m elev.), the season is less than 8 months Saguaro National 1976 12832 long. Summerhaven apiary at 2438 m elevation has a season Monument East of only 6.5 months. 11674 At the beginning of the flowering season, colonies at two of Falcon Valley 1980 -81 the apiaries, Tanque Verde and Summerhaven, experienced Summerhaven 1980 -81 22000 180 N 160: 160'

120. 120

C -r A L D D 1 _r ti 40: ti- 40; _r -C 1 J. Fm ì1 Ai S Ó No o N D

160 V 1 w 120 X

. 1. . -680 u E

40 \El

1 F M 1 1 A S 0 Ñ D

Figure 2. Profiles of the mean pollen influx per colony at five apiaries during 1976 near Tucson, Arizona. A = Steampump; B = Cortaro Road; C = San Xavier Mission; D = Tanque Verde; E = Saguaro National Monument East. Letters are above those samples of the profile that were analyzed. Results of the analyses are presented in Table 4. BIackened portions of the profiles are the contributions of plant taxa which did not 1U la present nectar to foraging bees and were thought to be more

F M Á M 1 1 Á Ó Ñ 0 or less anemophilous.

the Honey Bee was used to monitor flowering of vegetation. woodland elements in the bees' diet. Salix was both a major Occasionally, common plants were not included in the bees' riparian plant along the Santa Cruz River and the major pollen harvest when they flowered. Such exceptions included pollen source nearby at the Cortaro Road apiary. Creo- Ocotillo (Fougieria splendens) and members of the Milkweed sotebush (Larrea) in April and Mesquite (Prosopis), Saguaro Family (Asclepiadaceae), the pollen of which was not col- (Carnegiea), Palo Verde (Cercidium) and Ironwood (Olneya) in lected by Honey Bees, and certain anemophilous plants: May and June were all well represented in the pollen samples Wooly Plantain (Plantago insularis) in the Sonoran Desert at the apiaries in proximity to the associations that contained communities and Fir (Abies) and Douglas Fir (Pseudotsuga) these plants. at the Summerhaven apiary. In the autumn, the Sunflower Family (Compositae) repre- Near all the apiaries (except Summerhaven) are many sented a significant source of pollen for all of the apiaries legume trees parasitized by Desert Mistletoe (Phoradendron except Ragged Top. Autumn -flowering composites were rare Cali f ornicum). Pollen from P. calf f ornicum was the first pollen near Ragged Top. The low pollen influx at the Ragged Top of the year to be harvested at each of these apiaries and was apiaries in autumn, compared to the other apiaries, was the predominant pollen source during February at Ragged indicative of their absence. Burroweed (Haplopappus spp), Top. Jojoba (Simmondsia chinensis) was a major pollen source Telegraph Plant (Heterotheca subaxillaris), Snakeweed (Gutier- during March at Saguaro National Monument East, Ragged rezia spp), and Desert Broom (Baccharis sarothroides) were Top, and the Arizona -Sonora Desert Museum, where it was the major contributors for the desert apiaries. Summerhaven locally abundant. Cottonwood (Populus), Ash (Fraxinus), apiary samples in the autumn were composed predominantly Willow (Salix) and Walnut (Juglans) were important pollen of Compositae but, with the exception of Turpentine Bush sources at Tanque Verde apiary. Together they contributed a (Haplopappus larici f olius), the contributing species were dif- mean of 8.8% of the pollen to the Tanque Verde colonies ferent from those for the desert apiaries. In addition to the each year, a moderate but persistent representation of riparian composites, Wild Buckwheat (Eriogonum) was a major source Figure 3. Profiles of the mean pollen influx per colony of 8 apiaries at 6 different elevations near Tucson, Arizona. Vertical bars on a, b, and d indicate = 1 standard deviation around the mean of the variance of influx between colonies at the 500- same apiary. Vertical bars on e indicate = i standard deviation around mean of variance of the influx between 5 years. a = Summerhaven, 1980 -1981; b= Falcon Valley, 1980 -1981; e = 400 - Robles Junction, 1980 -1981; d = Arizona- Sonora Desert Museum, 1980 -1981; e = Tanque Verde, 1976 -1980; f = Painful, 1980 -1981; g = Harris Hawk, 1980 -1981; h = El Paso 300- Gasline, 1980 -1981.

200- of pollen at the Falcon Valley apiary in autumn. Falcon Valley and Robles Junction apiaries often had pollen samples low in pollen diversity compared with samples from the other apiaries. Such poor pollen diversity appeared to 100 reflect a similar low species diversity in the surrounding floral landscape. As mentioned previously, both of the apiaries in question are located adjacent to depauperate vegetation a, 2438 m resulting from overgrazing by cattle. Each pollen sample collected from the Tanque Verde apiary during 1976, 1978, 1979 and the first 5 months of 1980 was 100 analyzed to determine the species composition. This per- mitted presentation of individual influxes of 64 pollen taxa in Figure 4 and the calculation of their annual contributions (P) b, 1127m in Table 6. Samples collected during 1977 were not analyzed because the colonies were not closely managed during that year; bees through much of that season used holes in the hive ioo boxes that allowed them to by -pass the pollen traps. This + resulted in a reduced pollen influx. In Figure 4, monthly precipitation from 1976 to 1979 c, 820 m (recorded at Saguaro National Monument East Headquarters) and the mean monthly temperature are presented in the style c used by Simpson (1977). Mean monthly temperature values c too varied little from year to year in comparison with the u variability of total monthly precipitation. The overlay of these climatic parameters in Figure 4 is introduced to give a E d,792 m qualitative indication of the moisture availability for plants. When monthly precipitation lines fall well below the mean 100- monthly temperature line, plants experience water stress. In contrast, during winter (November- March) and summer (July - September) when the precipitation line exceeds the temper- e, 790m ature line, plants are relatively free from the inhibition of , water stress and vegetative growth can be expected. E 200 The total influx of pollen into the colonies at the Tanque Verde apiary (Figure 4) generally followed the bimodal á pattern of water availability in the Sonoran Desert. There = 100 were two major periods of pollen influx at Tanque Verde, 1) á February-June, and 2) August -October. Each period of pollen influx lagged about two to three months behind the periods 700m of high moisture. Fortuitously, the amount and pattern of rainfall for the goo early -spring flowering seasons of both 1978 and 1979 were unusually high, as occurs once every 30 to 40 years (Mc- Ginnies, 1980). The result was a profusion of growth and flowering of spring ephemeral herbs across the desert floor. In contrast, the spring flowering of 1976 was poor owing to scant rainfall and sparse growth of ephemerals. In 1976, there ioo were no pollen records for Twist Flower (Streptanthus), Fiddleneck (Amsinckia), and Scorpionweed (Phacelia) at Tanque Verde. Lupine (Lupinus) and Mexican Poppy (Esch- scholtzia), common . ephemerals throughout much of the Sonoran Desert in Arizona, were always uncommon near TimeofYear Elevation TIME PRECIPITATION mm TOTAL INFLUX POLLEN FLORAL TOTAL PELLET POLLEN INFLUX FROM VARIOUS TAXA, OF MONTHLY TOTALS AT OF POLLEN, R DIVERSITY, H' COMMUNITY BROOD DRY WT- COOL WINTER FLORA SAGUARO NATIONAL GRAMSCOLON9 I-DAY_1 H' mmZ ookmy -' ma xvs YEAR MONUMENT EAST HO p1loyoi N200 ACTIVITY, R-H' PERENNIAL `4 2 220 0 000 p0000 00 00 0 0 0 2 0 00o ç n 0-0 000 Qo o o R_ - 2,7422228:222, N -

1 1 1 I I I I 1 I I I I 1 nug I I ipJ.0,..t0, , I I limi IL1LI I1 il r,,,Iuul ln,,I .U,3".71LilJimitiÌ 1l ,l tW.L 1,14 ::, ,", , ÑJ hH

y DECEMBER n ::j i } - - i ------Ñ -1

Y i NOVEMBER _ .}. -- ._ i - - - I -I OCTOBER ',_4.74_ _. (¡ _

11 SEPTEMBER ' H N -i - ,n J AUGUST , _ `

. h ,1'

In JULY ._ . ` _ - - Ñ - --1 - J - - ' -- JUNE .,.. - -- - _ ; , . - - ' .- m -I N te 1 J L

.. i MAY _ i

. KS . APRIL _- -T _ _ - i ._. _:..- [ - Ñ-

O-I ....

i -.... - ._.i . , -..._ i - ir ___ , _ MARCH ' - i ...... t - -- - " -- - _ _1 l fl i '` _! ¡ _ rr_ D n FEBRUARY - .. .._.., .J4.- _... r.1- - Á -x- ' 4.

JANUARY v, - Á ' ¡ l'.1 I I I ( a e e e 3 IYnNNYi0 x rig-,°I -ñ° alÑ--g4R apta q+++ r- -ei n ñ 2p ñ á 2 II IIYII- H 6L61a'BL6199L610 .3anlva3aw3i áx mm sx«n<_Ma.^B° ^x «^O °J

FIGURE 4 :POLLEN DIAGRAM. UTILIZATION OF POLLEN OF THE SONORAN DESERT BY THE HONEYBEE, API$MELLIFERA L., BASED ON CORBICULAR POLLEN PELLETS COLLECTED FROM COLONIES WITH MODIFED O. A. C. POLLEN TRAPS AT TANQUE VERDE APIARY (Lot 32.14'N, Long.110.44m'W; Top I4S, RIGE, 5.6, SE /4 of NW /47 BESIDE TANQUE VERDE CREEK, PIMA COUNTY, ARIZONA; ELEVATION 790m.

LEGEND: - 1978 1978 __-__1977 ____ 1979

Tanque Verde and their pollen was rarely encountered in the tivity in arid lands range from 20.6 to 75 mm.year' (Noy -Meir, pollen traps at the Tanque Verde apiary. The yields per colony 1973; Seely, 1978). It is understandable that the value of 76 of spring ephemeral pollen during each of the two exceptional- mm should be at the high range because the yield is expressed ly wet seasons were 50 times greater than the yield during the in grams of pollen.colony', not grams of plant.meter'. Pollen is dry season of 1976 (Table 7). The linear regression of spring produced towards the end of the ephermerals' life, after some ephemeral pollen yield at the Tanque Verde apiary in plants have been consumed by herbivores (Inouye et al., 1980) g.colony' (Y) on total precipitation in mm during November - and after some have died from water stress. Also, not all of the April (X), Y = -1662.3 + 21.9X, has the high correlation coef- pollen produced by ephemerals is harvested by bees, and ficient of .98. This is not surprising since it is known that Honey Bees may ignore low density patches of flowers water is the major limiting factor for vegetative growth in (Schaffer et al., 1979). These factors would increase the zero - deserts and that the productivity of arid ecosystems is highly yield intercept value. correlated with rainfall (Noy -Meir, 1973). With the zero -yield Considering the linear regressions for each of the six intercept at 76 mm, the equation suggests that a threshold ephemerals, the threshhold amounts of precipitation for the amount of 76 mm of total winter precipitation must be ex- three ephemerals that typically grew on open ground (46 mm ceeded before pollen would be harvested from the six spring for Erodium, 78 mm for Lesquerella, and 82 mm for Strep- ephemerals. Known zero -yield intercepts for vegetative produc- tanthus) were less than the threshhold amounts of precipi- GRAMS COLONY' DAY WARM SPRING FLORA SuMMERFLORA COMPOSITAE EPHEMERAL oo 00=0000000200 0000000000 0000000 V229 222 2,22 &° 22é22° 22° 22é22°01°-mlf°IÓ2Ñ2 .7222 ONmAIOHOgN-

I 1 I I I 1, III,L 1 1 1 1 1 LIaw1,_,.1.h1,1. I,1,1,1,1,p1hi111 1 1,1J 1 I I 1,1,1,1,1, WIWI, ,I,I,I,,IJ, I Lhhh1. 1,1,111, 11,111,1,1,

R °2°RS% o ri6r h 22á ñ.°V:

ni sa - 4 var. a masi

JQ<2 J I CO ii 0 W 3 óó

tation for the three ephemerals that grew beneath shrubs or deeper to tap more remote reservoirs of ground water trees (91 mm for Sisymbrium, 95 mm for Amsinckia, and 123 (Cercidium, Acacia, Olneya, and Prosopis). These investments mm for Phacelia). One explanation is suggested: compared to reduced their phenologic dependence on the immediate the ephemerals associated with shrubs, open ground ephem- rainfall regimes. Flowering of these species was more depend- erals were in less competition with perennial plants for able. All profiles in this study from the Sonoran Desert had a available soil moisture and were likely to develop to the high influx of pollen during May and June. The pollen was flowering stage with less precipitation as a consequence. from these large dominant perennial plants. During growth, the ephemerals seemed the most sensitive An acute pollen dearth regularly occurred during the last of the desert life -forms to immediate rainfall- temperature week of June and the first weeks of July before the summer conditions. Their roots were shallow and a large proportion of rains. This dearth is concurrent with the warmest weather of the energy they acquired was soon transformed into seeds the year when intense heating of the desert valleys occurs before the plants died in the heat and drought of late spring. (Green and Sellers, 1964: 29). During and after the summer At the other end of the life -form continuum were the perennial rains, Pig Weed (Amaranthus), members of the Goosefoot arboreal forms including both columnar cacti and legume Family (Chenopodiaceae( and autumnal composites are the trees. These invested much of their energy into tissue either major sources of pollen. A second dearth exists from Novem- for storage of water (Opuntia and Carnegiea) or for rooting ber through January. The two dearths divide the year into two 92 Desert Plants 6 (2) 1984

Table 4. Analyses of selected samples of the 1976 pollen profiles Table 4, from five apiaries near Tucson, Arizona. Underlined taxa are those continued without nectar available in the inflorescences for foraging Honey Bees. N =200. Refer to Appendix I for a key to species and Figure 1 for TV N February 7 -9Cottonwood 98 %, Eucalyptus 2% abbreviations for apiaries. O April 3-7 Creosotebush 84 %, Eucalyptus 6.5 %, Sample and Rosaceae 3.5 %, Alfiliaria 1 %, Sow Apiary Time Span Composition Thistle 1 %, Elderberry 1 %, Globe Mallow 1 %, Tamarix .5%, Walnut .5%, SP A March 23 -24 London Rocket 33 %, Alfilaria 16 %, Mistletoe .5 %, Mulberry .5% Ragweed 14.5 %, Mistletoe 13.5 %, P May 15 -17 Mesquite 56.5 %, Creosotebush 30.5 %, Phacelia 6.5 %, Mulberry 5%, Eucalyptus Eucalyptus 6.5 %, Prickly Poppy 2 %, 4 %, Bladderpod 2 %, Fidddle -neck 1.5 %, Saguaro 2 %, Aloe 1 %, Bird -of- paradise California Poppy 1.5 %, legume 1 %, Aloe .5 %, Stick -leaf .5%, Bindweed .5% 1%, composite .5% Q May 29- Saguaro 44%, Mesquite 39%, Prickly B May 25 -28 Saguaro 70 %, Alfalfa 9%, Mesquite 7.5 %, June 2 Poppy 6.5%, Eucalyptus 5.5 %, Paloverde composite (4 spp.) 5.5 %, Eucalyptus 3 %, 1.5 %, Bird -of- paradise 1%, White -thorn White -thorn Acacia 1 %, Cholla 1 %, Acacia 1 %, Sow Thistle .5 %, Privet .5 %, Paloverde 1%, Tamarix 1 %, Stick -leaf unk. (1 sp.) .5% .5 %, Creosotebush .5% R July 31- C August 12 -13 Mesquite 41 %, Desert Marigold 30.5 %, Creosotebush 81%, Mesquite 6.5 %, August 2 Ceniza 5 %, White -thorn Acacia 3 %, Amaranth 8 %, Spiderling 5.5 %, Cheno - Sunflower 1.5 %, Peppergrass 1.5 %, am 3%, Alfalfa 2 %, Desert Zinnia 1.5 %, Sorghum .5 %, Bindweed .5 %, unk. (1 sp.) Sunflower 1.5%, unk. (1 sp.) 1.5 %, .5% Creosotebush 1%, Hackberry 1 %, Stick - leaf 1 %, Evolvulus .5 %, Burro -weed .5%, S October 6 -7 Burroweed and Telegraph plant 32%, Bird -of- Paradise .5 %, Hymenothrix .5 %, cheno -am 28 %, Desert Broom 27 %, Paloverde .5% Ragweed 9.5%, Wolfberry 1.5 %, grass .5 %, Puncture Vine .5 %, Sunflower .5 %, D September 3- 9Amaranth 65 %, Paperflower 11 %, unk. (1 sp.) .5% Spiderling 9.5 %, Stick -leaf 3 %, unk. (1 sp.) 2.5 %, Datura 2.5 %, Palm 1.5 %, Burro - SNME T March 16 -19Jojoba 91 %, Ragweed 5.5%, Manzanita weed 1.5 %, White -thorn Acacia 1%, 2 %, Mulberry .5 %, Joint -fir .5%, Hymenothrix 1%, Puncture Vine .5 %, Compositae 1 sp..5% Spurge .5%, Globe Mallow .5% U April 20 -23 Creosotebush 81.5 %, Walnut 8 %, E October 1 -4 Bindweed 70.5%, Burro -weed 13.5 %, Tamarix 7%, Globe Mallow 1 %, Palo- Desert Broom 8.5 %, Ragweed 3.5 %, verde .5 %, California Poppy .5%, Indigo -bush 2 %, Paperflower 1 %, Elderberry .5 %, unk. (2 spp.) 1% Amaranth .5%, Wolfberry .5% June 9 -11 Saguaro 86.5 %, White -thorn Acacia 6.5 %, CR F March 20 -22Willow 93.5 %, Alfiliaria 2 %, Cholla 5 %, Evolvulus 1%, Eucalyptus Creosotebush 1.5 %, Pine 1 %, Mulberry .5%, Compositae .5% .5 %, Ragweed .5 %, composite sp..5%, legume sp..5% W August 7 -9 Creosotebush 85 %, Evolvulus 5 %, Mesquite 2.5%, Sunflower 2.5 %,Spurge G April 23 -30 Mesquite 48.5%, Creosotebush 35 %, 1.5 %, unk. (1 sp.) 1.5%, White -thorn Tamarix 13.5 %, Paloverde 2 %, Cholla Acacia .5 %, Cactaceae .5 %, Ceniza .5 %, .5%, Rosaceae .5% Compositae .5% H June 3 -8 Saguaro 66.5 %, Mesquite 25 %, White - X October 6 -12 Desert Broom 52%, Burro -weed and thorn Acacia 2.5%, Eucalyptus 2 %, Telegraph Plant 30%, Jojoba 11.5%, Leguminosae 1.5%, grass .5%, Privet .5 %, Bindweed 3 %, Sunflower 2.5 %, Tamarix Tamarix .5 %, Mimosa .5 %, unk. 11 sp.) .5 %, Amaranth .5% .5% SXM I March 16 -19 Mulberry 63.5 %, Pine 28 %, Willow 2.5 %, Ash 2 %, Rosaceae 1.5 %, London Rocket discrete seasons of pollen availability. If the native zoophilous 1%, Eucalyptus .5 %, Globe Mallow .5 %, desert plants that are major pollen sources for the Honey Bee Leguminosae .5% are considered for both seasons, generally the larger the J. April 29- Mesquite 80.5%, Creosotebush 11.5 %, biomass of the plant, the later in the season it paticipates in March 5 Tamarix 7 %, Hackberry .5 %, Cruciferae the floral community and the more reliable its participation .5% becomes (Figure 5). A similar pattern of large life -forms flower- K June 9 -11 Saguaro 56%, Mesquite 32 %, ing later in the season than small life -forms is encountered Leguminosae 4.5 %, Privet 3.5 %, also in the northern prairie of the continent. Species that Eucalyptus 2 %, Prickly Poppy 1.5%, bloom there in May, June, July, August and September average Tobacco .5% 13, 23, 32, 34, and 46 cm in height, respectively (Butler, 1954). L July 31- Creosotebush 94 %, Ceniza 3 %, Tamarix But in northern deciduous forests, a reversal of life -form August 2 1 %, White -thorn Acacia 1 %, Mesquite .5 %, Sunflower .5% sequence has been noted, with bees harvesting pollen pre- dominantly from trees in early April to late May, then shrubs M October 6 -8Cheno -am 55.5 %, Ragweed 9.5%, in late May to mid -June, followed by herbs from mid -June Burro -weed 8.5%, Mesquite 7.5 %, Desert Broom 6%, Sorghum 6 %, Brittle -bush 3 %, through September (Synge, 1947; Severson and Parry, 1981). Leguminosae 2 %, Corn 1 %, Compositae The coefficient of variation of the annual contributions for 2 spp. 1% the three years by each of the 64 pollen taxa to the colonies at O'Neal and Waller Honey Bee Pollen Harvest 93

Honey Bee colonies with pollen traps at the Saguaro National Monument East apiary.

Tanque Verde apiary is presented in Table 6. This coefficient probability of interspecific encounter, o',, is perhaps more gives some indication of the reliability of the pollen harvest applicable as a parameter for measuring the potential resour- between years. The lower the coefficient, the greater the ces available to a Honey Bee feeding on the incoming supply reliability. For example, the mean coefficient for the six winter of pollen in a nest.) This graph only serves as a general zoophilous ephemerals (X = 122) is more than twice the mean indication of the activity of the floral community. When coefficient for the 7 large native and zoophilous trees and compared to the "Total Influx of Pollen," the graph does have shrubs (X = 57). Of the 10 zoophilous plants with coefficients the advantage that it reduces the representation of large of variation less than 50, 4 are a columnar cactus (Carnegiea) influxes produced by only one plant, such as Populus in early or large shrubs or trees (Larrea, Prosopis and Acacia). Four are February, and enhances large influxes contributed by several cultivated exotics, buffered from precipitation regimes by plants. The graph "Floral Community Activity" may be a good either irrigation (Leucophyllum, Acanthaceae, Caesalpinia) or representation of the nutritional well -being of the colony. by succulence (Aloe). The last two, Phoradendron calf f ornicum Gary et al.(1972) suggested that colonies increased the and Baccharis sarothroides, are notable: P cali f ornicum is a variety of pollen types harvested in order to increase the semi -succulent parasite that lives on desert legume trees probability of obtaining a nutritionally adequate diet. (Acacia, Prosopis, Olneya and Cercidium) and depends on its The nutritional value of pollen from different plant taxa hosts to supply a reliable source of water. Baccharis saro- should not be assumed equivalent for bee nutrition. Bio- throides is much like the large desert trees and shrubs that chemical analyses of pollen demonstrate large compositional bloom at the end of the first season of pollen availability. In differences between taxa (Stanley and Linskens,1974). the second (autumnal) season of pollen availability, B. saro- Maurizio (1960) arranged 35 plant taxa into categories throides is the largest plant, the more reliable of the flowering according to the degree of development of pharyngeal glands plants, and blooms at the end of the season. of Honey Bees raised on pollen from each given plant species. Dietary Diversity. The graph entitled "Floral Commmunity Entomophilous taxa were shown by this study to generally Activity" in Figure 4 presents qualitatively both the amount possess "good" pollen; most anemophilous angiosperm taxa of pollen harvested by the colonies at Tanque Verde and the "intermediate" pollen; and conifers "inferior" pollen. Pollen diversity (H') that the pollen harvest contains. It is appreciated bioassays that have included Dandelion (Taraxacum offi- that the diversity index, H', has several problems when cinale) and Fremont Cottonwood (Populus fremontii) indicated applied to the pollen data (see Hurlbert, 1971). (Hurlbert's that the pollen from these two plants was nutritionally 94 Desert Plants 6 (2) 1984

inadequate for the Honey Bee (Herbert et al., 1970; Loper and Because foraging Honey Bees filter-feed (by the action of Berdel, 1980). Pollen from some plants has been reported the proventricular valves) on ingested pollen from their toxic to Honey Bees when eaten alone. This includes pollen nectar loads (Maurizio, 1949; Bailey, 1952; Soehngen and Jay, from Monkshood (Aconitum spp) (Koptev, 1948; Poltev, 1956/, 1972) and individual foragers usually specialize on one plant Buckeye (Aesculus spp.) (Maurizio, 1945a; Vansell,1926), species (Brittain and Newton, 1933), one might expect Apis to Foxglove (Digitalis purpurea) (Muck, 1936), Buttercup (Ranun- avoid taxa with nutritionally inadequate or toxic pollen. culus spp.) (Maurizio, 1941, 1945b; Muller, 1948), and Death However, as corbicular pollen loads are deposited by the Camas (Zigadenus venenosus) (Hitchcock, 1959). foragers collectively in cells, the monolectic forager has access to the larger spectrum of pollen types harvested by the Table 5. Analyses of samples of the pollen influx from April 1 to colony. The diversity of pollen usually present in the pollen June 22, 1981, at Summerhaven apiary, Santa Catalina Mountains. stores probably gives the colony some degree of safety from Underlined taxa are anemophilous. N=200 pellets. poisoning or nutritional imbalances that could result from Influx Rate Time Span Composition harvesting and eating a single pollen type. (g col day) The Importance of Anemophilous Pollen in the Diet of the Honey Bee. Observations are frequently 100 April 1-19 Salix 95.5%, Acer 4.5% made of Honey Bees collecting pollen from a wide range of 263 April 20-22 Acer 8(15%, Salix 17%, Taraxacum anemophilous taxa (Schmalzel, 1980) in spite of the lower 2.5% expected nutritional value of anemophilous pollen compared 577 April 23-27 Acer 99%, Taraxacum .5%, unk. (I to zoophilous pollen. Also, it is often recognized that during sp.( .5% brief periods anemophilous plants contribute a significant 433 April 28-May 4 Acer 98.5%, Taraxacum 1%, portion to the pollen diets of Honey Bee colonies (Lovell, Rosaceae .5% 1926; Louveaux, 1958, 1959; Riedel and Wilson, 1967; Olsen, 276 May 5-11 Acer 95.5%, Taraxacum 3.5%, 1975), Pollens from taxa such as Date Palm (Phoenix dactyli- Cyperaceae .5%, unk. (1 sp.) .5% fera), Corn (Zea mays), and Beech (Fagus sylvatica) are con- 87 May 12-14 Acer 72.5%, Quercus 4.5%, sidered major constituents of the annual diet of Honey Bee Taraxacum 2.5%, Salix 2%, colonies at some locations in the world (Ibrahim, 1976; Lupinus 1.5%, Rosaceae 1%, Ribes McLellan, 1977). Yet there is no discussion in the literature 1%, unk. (2 spp.) 15% about what the implications may be if Honey Bees are reliant 50 May 15-18 Taraxacum 17%, Robinia 9.5%, on anemophilous pollen for much of their diet. Salix 7%, Acer 5.5%, Lupinus The proportion of anemophilous pollen (and zoophilous pol- 4.5%, Quercus 4%, Compositae (1 sp.) 1%, Ribes .5%, unk. (I sp.) len from species lacking nectar available to the bee) in each 51% analyzed sample from the 5 apiaries in 1976 is indicated by 54 May 19-26 Quercus 36%, Lupinus 9%, black bars in Figure 2. Anemophilous plants at these apiaries Robinia 4%, Compositae (1sp.) between 730 and 950 m elevation were major sources of pollen 2.5%, Taraxacum 2%, Salix 1%, for the Honey Bee during the February-March and September- Dasylirion 1%, Ribes .5%, October intervals. These are the same intervals of maximum Cyperaceae .5cYo, unk. (1 sp.) 43.5% anemophilous pollen production in the Tucson area (M. 53 May 27-June 1Quercus 20.5%, Lupinus 17%, O'Rourke, personal communication). The anemophilous pol- Robinia 4.5%, Taraxacum .5%, len was harvested from a number of plants including Mulberry unk. (1 sp.) 57.5% (Morus), Pig Weed (Amaranthus), members of the Goosefoot 68 June 2-8 Gramineae 115%, Heracleum Family (Chenopodiaceae), Jojoba (Simmondsia), Pine (Pinos), 19.5%, Rubus 13%, Quercus Cottonwood (Populos), grasses (Gramineae), and the Ragweed 10.5%, Lupinus 9.5%, Cornus 3%, Phragmidium (rust) 1%, Tribe (Ambrosiinae). Since all samples were analyzed from Compositae (2 spp.) 196, Robinia Tanque Verde apiary for 3 years and the annual contribution .5%, unk. (3 spp.) 18.5% for each plant calculated in Table 6, the proportion of 12 June 9-12 Quercus 20%, Lupinus 10.5%, anemophilous pollen to the annual harvest was determined Rubus 2%, Cornus 1.5%, Robinia and is presented in Table 8. Anemophilous pollen contributed 1.5%, Compositae (2 spp.) 2%, from 9 to 24% of the annual pollen harvest during the 3 years. Dasylirion 1%, Gramineaell Heracleum .5%, Phragmidium The major anemophilous sources were Cottonwood (Populusl, .5%, Pinus unk. (1 sp.) 59% Pig Weed (Amaranthus), grasses (Gramineae) and members of the Ragweed Tribe (Ambrosiinae). 13 June 13-15 Heracleum 27.5%, Rubus 20%, Arceuthobium 19.5%, Grainineae All of the plant associations available to the colonies at the 8%, Lupinus 7.5%, fuglans 2.5%, top of the Santa Catalina Mountains (Summerhaven apiary) Rumex 2%, Phragmidium 1%, were dominated by anemophilous plants. Only meager Quercus .5%, Dasylirion .5%, amounts of nectar, primarily from Sweet Clover (Melilotus) Caryophyllaceac .5%, unk. (2 spp.) and Cow Parsnip (Heracleum) were available to the bees. The 10.5% colonies did not have more than 1 kg of fresh, unsealed honey 30 June 16-22 Heracleum 36.5%, Gramineae 33%, Robos 10.5%, Arceuthobium Pinus .5%, Caryophyllaceae .5%, Rumex .5%, unk. ¡3 spp.) continued on 12.5% page 99 O'Neal and Waller Honey Bee Pollen Harvest 99

rarely been done and plant host records for bee species are Pollen Harvest undoubtedly biased towards melittophilous plants since col- continued from lectors generally pursue bees on "bee flowers." Nevertheless, page 94 the occasional observations of insect visitors to anemophilous in their combs at any time during the summers. These bees plants indicates how common it may be for solitary bees to were dependent for their survival upon honey given to them exploit this pollen. A Golletes sp. has been observed foraging in combs); each colony received and consumed ca 70 kg of in high numbers for pollen on Juniperus (Burnham, 1978). honey between July, 1980 and July, 1981. Because of the Female Anthophora neglecta Timberlake and Cockerell have nature of the surrounding vegetation, it was anticipated that been observed collecting pollen from Arizona Sycamore anemophilous pollen would be very important in the colonies' (Platanus wrightii) (O'Neal, personal observation). Some diet. Based on the analyses of the pollen collected in the traps pollen loads in the scopae of individuals of Andrena haemor- on the Summerhaven colonies from April 1 to June 22, 1981 rhoa Fabr., A. armata Gmelin and A. pubescens Oliv. in (Table 5), it was determined that 9.00 kg or 71% of the 12.66 England were composed of more than 50% Beech (Fagus) or kg (X of all colonies) of pollen collected during this period Oak (Quercus) (Chambers, 1945). The Catalog of Hymenoptera were from Acer and Quercus. Even if the Honey Bees in America North of Mexico (Krombein et al., 1979) lists a included no anemophilous pollen in their diet after June 22, total of 22 anemophilous plant genera in the visitation records since 22 kg of pollen per colony was estimated as the total of 59 species (17 genera) of bees. These 59 species represent yield for 1981, the anemophilous pollen collected prior to only 1.7% of the bee species in America north of Mexico. June 22 represented at least 40% of their annual diet. Among the bee, Apis and the meliponines, with their large, Analyses of pollen diets of Honey Bees in other regions perennial honey stores, ought to be most capable of securing indicated a wide range of dependency on anemophilous anemophilous pollen. Except for a few field observations pollen. Severson and Parry (1981) showed that colonies and /or nest pollen analyses of Trigona and Melipona (Turner, during one season in southwestern Wisconsin relied on 1974; Absy and Kerr, 1977; Engel and Dingemans -Bakels, anemophilous plants, especially Oak (Quercus) and Corn 1980 and references therein), pollen diets of meliponine bees (Zea), for more than 40% of their pollen during 63 out of the remain poorly known. The initial dietary surveys suggest 159 foraging days available to the bees. In a Kansas flood that anemophilous pollen sources are of minor importance plain, 18.2% of the annual pollen diet was derived from for most meliponine bees compared to Apis mellifera. anemophilous plants (Rashad and Parker, 1956). In Switzer- Bumblebees (Bombus), unable to sequester large caloric land, anemophilous pollen contributed approximately 56% stores in their colonies (Heinrich, 1979), appear to be under and 23% in 1950 and 35% and 33% in 1951 of the annual diet energetic constraints to harvest pollen closely associated for colonies at Liebefeld and Wadenswil, respectively (Mauri- with nectar. In this respect, Bombus is more like solitary bees zio, 1953). In some regions, Honey Bees rely relatively less on than highly social bees. Adams et al. (1979) observed Bumble- anemophilous pollen. A colony studied by Percival (1947) in bees that were foraging primarily in patches of melittophilous Wales relied on anemophilous pollen for only 0.5% of its flowers but made occasional trips to Plantago to collect annual diet. Based on the combined averages of two colonies pollen. The observation by Leuck and Burton (1966) of in England, anemophilous pollen contributed no more than Bombus collecting pollen from Pennisetum americanum 8.6% to the annual diet (Synge, 1947). does not mention the general availability of nectar at the Of the 0.01 to 10 metric tons.km 2 of pollen produced by locality. It would be interesting to investigate the importance the vegetated regions of the earth (Solomon, 1979), most is of anemophilous pollen for Bombus colonies in regions where produced by anemophilous plants. A bee species that can they utilize honeydew as a nectar substitute (see Maurizio, consistently exploit anemophilous pollen has a larger food 1964), since honeydew might represent a long -term carbohy- base available than a bee harvesting only zoophilous pollen. drate superabundance for Bombus. Knee and Moeller (1967), However, harvesting anemophilous pollen requires a reliable in their comparative study of pollen sources of Apis and source of nectar or a store of honey in order to periodically Bombus in the same field at the same time, found that the replenish either energy expended in flight or crop fluids used major sources of pollen for Bombus fervidus (Fabr.) were Red to moisten and manipulate the powdery anemophilous pollen Clover (Trifolium pratense) and Alfalfa (Medicago sativa). into the bee's scopae (Schremmer, 1967). Although Honey Bees included these two plants, their major At times, bees may discover nectar in much greater quan- pollen source was Corn (Zea mays). Bombus fervidus did not tities than is required to meet flight costs. This condition utilize Zea. should permit more freedom for bees to examine sources of The importance of anemophilous pollen in the diet of pollen not immediately associated with nectar. Yet, because solitary bees is indicated approximately by the proportion episodes of nectar superabundance in a floral community are contributed by anemophilous plant taxa to the nest provi- probably brief, patchy and unpredictable, a carbohydrate sions. But for social bees, the timing of the pollen harvests surplus is a condition that a species of solitary bee probably may be as important as the annual proportions contributed never experiences. Social bees, with their capacity to store by the various plants. This needs further examination. A carbohydrates in the nest for later use, are expected to utilize Honey Bee colony that harvests anemophilous pollen in large anemophilous pollen to a greater extent than those classifed quantities early in the season when zoophilous pollen is as solitary bees. unavailable probably increases the rate of brood production It is difficult at the present to assess the general importance relative to the rate of a colony that depends only on zoophi- of anemophilous pollen in the diet of solitary bees. The most lous pollen, thereby increasing the probability of colony objective means, analysis of pollen in the nest provisions, has reproduction during the season. Protective clothing is used when removing afullpollen drawer from a trap. Table 6. Net (g) and relative ( %) annual pollen contributions from variousplants to pollen traps on Honey Bee colonies at Tanque Verde apiary. Refer to Appendix 1 for key to species.

Mean Net Contributions3 Pollination Time of Relative Standard Coefficient Plant Mode' Harvest Contribution2 1976 1978 1979 Mean Deviation of Variation 56 Cottonwood A Feb. 5.4 1118 293 985 799 443 Eucalyptus Z Oct: July 1.8 648 1058 19 575 523 91 A March 2.1 404 148 408 320 149 47 Ash 42 Mistletoe Z Jan. -March 1.9 212 424 223 286 119 90 Willow A Z March 1.3 126 75 417 206 185 173 Rhus Z Jan. + 0 9 0 3 5 85 Walnut A April + 4 4 15 8 7 8 103 Elm A Feb. .1 7 0 15 8 73 Mulberry A March -April .1 26 4 34 21 16 142 Pine A March .2 63 0 9 24 34 2 3 138 Joint -fir A March + 6 1 0 173 Manzanita Z Feb. + 5 0 0 2 3 59 Plum Z Feb: March .2 17 26 53 32 19 139 Other Rosaceae Z Nov: April .2 55 <1 ? 28 38 71 Rosemary Z Nov. -March .1 3 19 23 15 11 17 Aloe Z Nov. -May .1 19 16 13 16 3 13 89 Globe- mallow Z Nov. -May .1 12 4 30 15 120 Dock A March -April + 0 4 14 6 7 135 Prickly Poppy Z April -July .1 30 0 6 12 16 59 Peppergrass Z April -Oct. 1.0 59 235 153 149 89

Winter Ephemerals 117 Bladderpod Z Feb.-April 4.5 7 1650 498 718 844 643 92 London Rocket Z Jan.-April 4.1 1 842 1265 702 112 Twist Flower Z March -April 1.4 0 511 181 230 259 93 Alfilaria Z Jan.-May 5.3 72 862 1731 888 830 148 Fiddleneck Z March -April 1.6 0 85 761 282 417 169 Phacelia Z March -April .1 0 1 58 20 33 for winterephemerals, X =122 Native Large Shrubsand Trees Z March -Oct. 10.4 2135 878 1534 1515 629 41 Creosotebush 134 Paloverde Z mostly May 1.5 95 576 10 227 305 Mesquite Z April -Oct. 14.0 1767 2361 2303 2144 328 15 350 41 Saguaro Z May -June 5.4 649 651 1257 852 66 Desert Willow Z May -August .1 5 14 24 14 9 White -thorn Acacia Z May -August .9 102 186 117 135 45 33 Cholla and Prickly Pear Z May .8 24 160 192 125 89 71 for native largeshrubs and trees, X= 57

Privet Z May -June .4 28 59 127 71 50 71 119 Bindweed Z May -Sept. .6 27 232 35 98 116 Z Sept-Feb. + 16 7 0 8 8 106 Wolfberry 26 Ceniza Z July -Oct. .9 172 101 138 137 36 1 10 Acanthus Z July -Sept. .1 11 10 12 11 1 1 92 Agave Z July- August + 1 1 0 12 69 Spiderling Z July -Oct. .1 21 4 28 18 4 27 Bird -of- Paradise Z April- August .1 20 11 16 16 3 2 76 Hackberry Z July- August + 1 4 5 72 162 Sotol Z June -August .3 <1 127 6 44 Datura Z July -Nov. + 4 + 10 5 5 105 135 Sangre -de -Cristo Z July -Oct. .1 0 10 56 22 30 5 90 Stick -leaf Z August -Sept. .1 7 0 10 6 Evening Primrose Z June -Sept. ..3 3 9 120 44 65 148 49 54 110 Clammyweed Z July -Sept. .3 11 ? 87 79 41 52 .5 - 116 86 Tamarix - Z April -Oct. 35 Puncture Vine Z July -Sept. .2 5 34 95 45 46 103 140 Corn A June -Sept. .2 8 86 5 33 46 114 Other grasses A June -Nov. .4 25 134 15 58 66 Amaranth A July -Oct. 6.0 1172 522 982 892 335 38 Other cheno -am A All year 1.1 222 227 55 168 98 58 All composites A Z All year 18.9 2557 1534 4786 2959 1663 56 65 Sow Thistle Z Jan.-July .6 79 38 154 90 59 Paperflower Z Anytime 2.8 76 468 833 459 379 83 Wild Lettuce Z June -July <.1 5 12 4 7 4 62 Sunflower Z July -Oct. .7 65 47 217 110 93 85 140 127 Hymenothrix Z August -Oct. .7 12 ? 210 111 Burroweed and Telegraph plant Z August -Oct. 9.6 1686 465 2231 1461 904 62 41 47 Ragweed A Sept-Oct. .6 134 73 57 88 Desert Broom Z Sept-Oct. 3.2 454 321 716 497 201 41

' A= anemophilous, Z= zoophilous a % of total yield/colony /year 3 P; grams /colony /year 102 Desert Plants 6 (2) 1984

Pollen -Harvesting by Honey Bees as a Measure ity, the pollen influx data from various regions of the world of Global Flowering. All profiles of the annual pollen can be compared to see what patterns of global floral availa- influx into Honey Bee colonies indicate that influx rates vary bility may exist. When the Arizona pollen profiles are compared almost continuously through time (apart from those rates with profiles from other latitudes, in seasonality and magni- equal to zero during winter or severe flowering dearths.) The tude, the Arizona profiles from high elevations resemble best representations of this continuous and occasionally those from higher latitudes; those from low elevations are like dramatic change in influx rates are profiles of the influx into those from adjacent latitudes. The profile at 2438 m elevation one colony when the time intervals are short (for example, (Figure 3a) is similar to the profile from 50 °N latitude in 1 -day intervals in Figure 6e). Despite averaging effects, even British Columbia (Figure 6a). Both last for only 6 months each profiles based on mean influxes between colonies, days year and both have a major pulse of influx during the first two and /or years show that influx rates are hardly static (Figures months of the season. As expected the lower elevations and 2, 3 and 6). latitudes have the longest harvesting season, approximately The fluctuating mode of pollen harvesting by Honey Bee 10 months at 30 °N and 12 months at 10 °N (Ramírez, 1980). colonies is possible not only because there may be many idle Longer seasons do not result in greater annual yields. The bees in the colony that can be rapidly recruited to profitable lowest reported annual yield comes from the tropics, 2.6 kg patches of flowers, but also because a successful forager can per colony at Alajuela, Costa Rica (Ramírez, 1980), and the quickly unload its pollen or nectar into the colony's nest and highest yield, from early successional vegetation in the mid - resume foraging. Thus, with both 1) bees communicating latitudes, 53.1 kg per colony at Davis, California (Eckert, information concerning profitable floral patches and 2) combs 1942). In Figure 6, the influxes of the profiles between 35° and in the colony functioning as extensions of the bees' stomachs, 50 °N latitude at least briefly exceed 200 g/colony /day each the colony is extremely responsive to sudden flushes of floral year. If the single data set of Ramírez (1980) is representative resources and capable of persisting in harvesting large quanti- of tropical environments, the high influxes in the temperate ties of a major pollen or nectar source. With this potential zones appear to be uncommon at low latitudes, since the responsiveness to fluctuations in floral resource availability, colonies of Ramírez failed to sustain even moderately good the Honey Bee colony is basically opportunistic in foraging rates of pollen harvesting (greater than 50 g/colony /day). and the pollen influx should be some function of actual floral More tropical data are needed to ascertain what factors are availability. limiting yields. Seasonality and Magnitude. If the pollen influx into Figure 6 suggests that at higher latitudes (45 -65 °) there is a a colony is assumed to be a good indication of floral availabil- unimodal pulse in pollen availability; at somewhat lower

FIRST FLORAL WAVE

increasing biomass,increasingreliability

m m E

FEBRUARY MARCH 'APRIL 'MAY JUNE JULY winter ephemerols: Erodium, Streptanthus, Larrea Opuntia legume trees:Prosopis, Cercidium, Encella Cornegiea Amsinckia, Phacelia, Lesquerella and Olney °, Acacia Sisymbrium O'Neal and Waller Honey Bee Pollen Harvest 103 latitudes (30 -45 °), there is a bimodal pulse in pollen availabil- The abundance -diversity curves of the data sets are corn- ity. On a global scale, floral resource availability appears to be pared with importance values based both on percent contri- a wave -form phenomenon in the temperate latitudes. Honey bution (Figure 7) and on absolute contribution (Figure 8). Bee colonies tend to swarm shortly after a major pulse in These abundance -diversity curves all describe log -normal floral availability. The global pattern of reproductive swarm- distributions. The curves illustrated are somewhat steeper ing (O'Neal, personal observation) supports the idea that than they would have been if the ranking were strictly based there is a major floral wave which completes one oscillation on species. Some of the ranked taxa are genera (e.g. Tri f olium, between poles each year, reaching the higher latitudes once Melilotus, and Cercidium) or are a mixed group (e.g. Acer- annually, but passing through the mid -latitudes twice an- Pyrus -Prunus) because of the difficulty in resolving pollen to nually. Since floral resource availability has not been quanti- the species level for some taxa. fied in the tropics, it is not known whether this floral wave Although the contrasts between plots on the graph are not persists or dampens out in the tropics. as great as expected, consistent differences exist. Plots of the Abundance -Diversity Relationships. Whittaker three years' data from Tanque Verde apiary (32 °N) are gener- (1965, 1972) and Hubbell (1979) have noted that different ally well above those of England and Wales (51 °N). The plant communities or different successional phases within Kansas plot (39 °N) lies in an intermediate position between the same community produce characteristic curves when the plots from farther north and south. species of plants are ranked from most to least in importance. The effect of increased rainfall during 1978 and 1979 at If many species are contained within the assemblage, as is the Tanque Verde best explains why the plots of these two years case in a tropical wet forest, a log -normal distribution of log are above that for the dry year of 1976. More plants contri- importance plotted against species rank is obtained. Com- buted greater proportions of pollen during the wet years than munities with few total species, such as a subalpine forest in during the dry year. The ten top- ranked species were not the which one or a few species comprise almost the entire plant same each of the years at Tanque Verde, nor were there more community, should produce a simple geometric series (May, of the top ten species shared between the two wet years (6 1975; Hubbell, 1979). species in both 1978 and 1979) than between the wet and dry Comparable data sets of vegetation are scarce, but when years (5 in both 1976 and 1978; 6 in both 1976 and 1979). available such plots are useful for descriptive analysis of Only four plants, Creosotebush (Larrea), Mesquite (Prosopis), abundance -diversity relationships of the assemblages. Annual Saguaro (Carnegiea), and Pig Weed (Amaranthus), ranked data sets on relative and absolute contributions of pollen by among the top ten contributors every year. plant species to a Honey Bee colony are directly comparable In relative proportions contributed, the rank -1 and rank -2 between localities across North America and most of Europe species of all three localities north of Tanque Verde consis- since the present populations of Honey Bees are intermixed. tently contributed more than the two top- ranked species at Regardless of the region from which it is obtained, a new Tanque Verde. In absolute quantities contributed, only the colony at an apiary can be expected to harvest pollen in a top two species at the Wales location were well above those at manner similar to that of colonies long established at that Tanque Verde. apiary. The diet of a Honey Bee colony with a geometric series - There are only four study localities where the diets of type abundance -diversity curve is more limited than the diet Honey Bee colonies have been analyzed thoroughly enough indicated by a log -normal distribution. Late freezes or in- to construct abundance -diversity curves. Two localities are in clement weather for foraging, for example, could significantly rural semiagricultural landscapes in England (Synge, 1947) reduce the contribution of pollen from one of the normally and Wales (Percival, 1947). One is in a Kansas river floodplain major- ranked species. The loss of a major- ranked species in (Rashad and Parker, 1956), and one, in the Sonoran Desert a geometric series diet will significantly reduce the total (Tanque Verde apiary). annual pollen harvest, whereas, with log -normal distributions,

Figure 5. Zoophilous species which SECOND FLORAL WAVE had larger biomass bloomed later in the flowering season. These larger plants increasing biomass, increasingreliability were more reliable as pollen sources from year to year than the smaller ones. Two distinct floral waves occured. The first, extending from February into July, began with small winter ephemerals passed to Brittlebush and Creosotebush, then to Prickly Pear and Saguaro and m finally to legume trees. The second, m E extending from July through October, began with low lying Zinnia and Paper - JULY AUGUST SEPTEMBER OCTOBER flower, passed to Sunflower and Tele- Hymenothrix Gutierrezia Zinnia Baileya Baccharis graph Plant and finally to Desert Broom. Happlopappus Psilostrophe Heterotheca Helianthus Figure 6. Profiles of the seasonal pollen harvest by Honey Bee colonies at various latitudes. a= Vernon, British Columbia; 1943, based on original data developed by W H. McMullen for Vansell and Todd (1949). b = Munich, Germany; X ± 1 standard deviation of 2 years' data; redrawn from Hirschfelder (1951). e = Corvallis -Albany, Oregon; 1949, based on original data developed for Vansell and Todd (1949). d = Guelph, Ontario; 1977, data used with permission from M. V. Smith a, 30.1 N (Univ. of Guelph, Canada). e = Manhattan, Kansas; 1954, 1955; Rashad and parker (1956). f = Davis, California; X ± 1 standard deviation of 2 years' data; redrawn from Eckert (1942). g = southwestern Arkansas; 1941 -1951; X ± 1 standard deviation of 11 years' data; Thompson (1960). h = Tanque b, 49.3N Verde, Arizona; X ± 1 standard deviation of 5 years' data; 1976 -1980. i = Baton Rouge, Louisiana; Whitcomb (1944) and original data developed by Kauffeld (1980).

the diet should be little affected. The curves from Kansas, Wales, and England indicate much greater species abundance- diversity than would be typical for vegetation at these latitudes unmanipulated by c, 44.3 N human activity. Without the effect of human disturbance on the vegetation, these curves would probably resemble more closely geometric series. By increasing the number of plant species potentially exploited by Honey Bees, humans have probably contributed significantly to the development and persistence of habitats suitable for bee populations. This process began in Eurasia at least several thousand years ago d, 43.3N and can be expected to accelerate throughout much of the world. Acknowledgements We thank Dr. Vera Markgraf- Bradbury and Dr. Robert Gilbertson for their assistance in identifying pollen and , 39.1N fungal spores and for reviewing the manuscript, Dr. Stephen c 400 Buchmann and Dr. Martha Gilliam for reviewing the manu- ó script. Joseph Martin for managing the colonies, Gail Lovitt, a Kelly Smith, Charles Shipman and Gene Joseph for collecting 300 pollen on numerous occasions, Dr. M. V. Smith for his pollen influx records and the late S. E. McGregor for the original

200 pollen influx data developed in the 1940's by the late E. E. Todd.

ioo

References f, 38.5N Absy, M. L. and W. E. Kerr. 1977. Algumas plantas visitadas para obtençáo de pólen por operárias de Melipona seminigra merrillae em Manaus. Acta Amazonica 7: 309 -315. ioo- Adams, R. J., G. C. Manville and J. H. McAndrews. 1978. Comparison of pollen collected by a honey bee colony with a modern wind - dispersed pollen assemblage. Canad. Field -Nat. 92: 359 -368. g, 33.4N Adams, R. J. and J. K. Morton. 1972, 1974, 1976, 1979. An Atlas of Pollen of the Trees and Shrubs of Eastern Canada and the Adjacent United States. Part I (1972): Gymnospermae to Fagaceae. 52 pp. Part II (1974): Ulmaceae to Rosaceae. 53 pp. Part III (1976): Leguminosae to Cornaceae. 37 pp. Part IV (1979): Clethraceae to Caprifoliaceae. 27 pp. Biology Series, Univ. Waterloo, Waterloo, Ontario. Adams, R. J., M. V. Smith and G. F. Townsend. 1979. Identification of h, 32.1N honey sources by pollen analysis of nectar from the hive. J. Apic. Res. 18 :292 -297. Anonymous (Rothamsted Exp. Sta.). 1947. Pollen collection by honey- bees. Bee Craft 29:25 -26. Bailey, L. 1952. The action of the proventriculus of the worker honey I, 30.3N bee Apis mellifera L. J. Exp. Biol. 29:310 -327. Bailey, L. H. 1973. Manual of Cultivated Plants. MacMillan Co. New Time of Year Latitud York. 1116 pp. Table 7. Total winter precipitation [X, mm) from November through April for four years at Saguaro National Monument East [8 km south of Tanque Verde) and the mean yields (Y, grams colony -1) of pollen harvested by colonies at Tanque Verde apiary from six springephemerals: Erodium, Lesquerella, Streptanthus, Sisymbrium, Amsinckia, and Phacelia.

Year 4-Year 1976 1978 1979 1980 E m b Y=0 r Precipitation 72 238 285 175 Pollen Yields Erodium 72 862 1731 1230 3895 6.7 -305 46 .88 Lesquerella 7 1650 498 5 2160 4.7 -369 78 .56 Streptanthus 0 511 181 0 692 1.6 -129 82 .60 91 .96* Sisymbrium 1 842 1265 281 2389 5.9 -539 Amsinkia 0 85 761 155 1001 2.8 -291 95 .75 123 .67 Phacelia 0 1 58 0 59 .2 -26 Total Yield 80 3951 4494 1671 10I96 21.9 -1660 76 .98* % Contribution by Ephemerals to Annual Yield .6 25.1 22.7 8.7 Statistics are for standard least squares linear regression equations: Y =mX + b; Y =0: the x- intercept ofthe calculated linear regression curve indicates minimum rainfall for pollen yield; r= correlation coefficient. * Above 98% level of significance with one -tailed test.

Bailey, L. H. 1976. Hortus Third: A Concise Dictionary of Plants Table 8. The contribution of pollen from anemophilous plants at Cultivated in the United States and Canada. MacMillan Publ. Co., Tanque Verde apiary. New York. 1290 pp. Betts, A. D. 1928. Pollen -supply, brood- rearing and the nectar flow. Year Yield of Anemophilous Proportion of anemophilous Pollen to Total Yield Bee World 9:23 -25. Pollen CX g/colony'year -i) Brittain, W. H. and D. E. Newton. 1933. A study in the relative 1976 3067 .241 constancy of hive bees and wild bees in pollen gathering. Can J. 1978 1419 .090 Res. 9:334 -349. 1979 2607 .132 Burnham, L. 1978. Survey of social insects in the fossil record. Psyche 85:85 -133. Butler, J. E. 1954. Interrelations of Autoecological Characteristics of Prairie Herbs. Ph.D. Diss. Univ. of Wisconsin. Madison. Chambers, V. H. 1945. British bees and wind -borne pollen. Nature Hirschfelder, H. 1951. Quantitative untersuchungen zum pollen - 155:145. eintragen der bienenvolker. Z. f. Bienenforschg. 1:67 -77. Chauvin, R. and J. Louveaux. 1956. Les caractères de la récolte de Hitchcock, J.D. 1959. Poisoning of honey bees by death camas pollen en 1955. La flore Francaise productrice de pollen pour les blossoms. Am. Bee. J. 99:418 -419. abeilles. L'Apiculteur 100:17 -28. Huang, T. 1972. Pollen Flora of Taiwan. Natl. Taiwan Univ. Bot. Dept. Crane, E. 1976. Bee products: harvesting pollen from hives. Bee World Press. Taiwan. 297 pp. 57:20 -25. Hubbell, S. P. 1979. Tree dispersion, abundance, and diversity in a Eckert, J. E. 1942. The pollen required by a colony of honeybees. J. tropical dry forest. Science 203:1299 -1309. Econ. Entomol. 35:309 -311. Hurlbert, S. 1971. The nonconcept of species diversity: a critique and Engel, M. S. andF. Dingemans -Bakels. 1980. Nectar and pollen alternative parameters. Ecology 52:577 -586. resources for stingless bees (Meliponinae, Hymenoptera) in Suri- Ibrahim, S. H. 1976. A list of pollen plants visted by honeybees in nam (South America). Apidologie 11:341 -350. Egypt. Agric. Res. Rev. (Cairo) 54:217 -219. Erdtman, G. 1952. Pollen Morphology and Plant Taxonomy. 1. Angio- Inouye, R. S., G. S. Byers and J. H. Brown. 1980. Effects of predation sperms. The Chronica Botanica Co. Waltham. 539 pp. and competition on survivorship, fecundity, and community struc- Free, J. B. and I. H. Williams. 1971. The effect of giving pollen and ture of desert annuals. Ecology 61:1344 -1351. pollen supplement to honeybee colonies on the amount of pollen Jebsen, W 1952. Das bienenleben in zahlen. Archiv für Bienenkunde collected. J. Apic. Res. 10:87 -90. 29:17 -30. von Frisch, K. 1967. The Dance Language and Orientation of Bees. Jeffree, E. P. and M. D. Allen. 1957. The annual cycle of pollen storage Harvard Univ. Press. Cambridge. by honey bees.J. Econ. Entomol. 50:211 -212. Gary, N.,P. Witherell and J. Marston. 1972. Foraging range and Kapp, R. O. 1969. How to Know Pollen and Spores. Wm. C. Brown Co. distribution of honey bees used for carrot and onion pollination. Dubuque. 249 pp. Environ. Entomol. 1:71 -78. Kauffeld, N. M. 1980. Chemical analysis of Louisiana pollen and Green, C. R. and W. D. Sellers. 1964. Arizona Climate. Univ. Arizona colony conditions during a year. Apidologie 11:47 -55. Press. Tucson. Kearney, T. H. and R. H. Peebles. 1960. Arizona Flora. Edition 2. Univ. Hare, Q. A. and G. H. Vansell. 1946. Pollen collection by honeybees in Calif. Press. Berkeley. 1085 pp. the Delta, Utah, alfalfa seed- producing areas. J. Amer. Soc. Agron. Knee, W. J. and E E. Moeller. 1967. Comparative study of pollen 38:462 -469. sources of honeybees and bumblebees. J. Agric. Res. 6:133 -138. Harris, W. E. and D. W Filmer. 1948. Pollen in honey and bee loads. Koptev, V. 1948. [The reasons for the summer deaths of bees in New Zeal. J. Sci. Technol. 20:178 -187. Siberia.] Pchelovodstvo 10:45 -47. (In Russian). Haydak, M. H. 1970. Honey bee nutrition. Ann. Rev. Entomol. Krombein, K. V., P. D. Hurd, D. R. Smith, B. D. Burks and others. 1979. 15:143 -156. Catalog of Hymenoptera in America North of Mexico. Vol. 2 Heinrich,B.1979. Bumblebee Economics. Harvard Univ. Press. Smithsonian Inst. Press, Wash., D.C. Cambridge. Lehr, J. H. 1978. A Catalogue of the Flora of Arizona. Desert Bot. Herbert, E., W Bickley and H. Shimanuki. 1970. The brood -rearing Garden. Phoenix. 203 pp. capability of caged honey bees fed dandelion and mixed pollen Leuck, D. B. and G. W. Burton. 1966. Pollination of pearl millet by diets. J. Econ. Entomol. 63:215 -218. insects. J. Econ. Entomol. 59:1308 -1309. Heusser, C. J.1971. Pollen and Spores of Chile. Modern Types of Loper, G. and R. Berdel. 1980. A nutritional bioassay of honeybee Pteridophytes, Gymnospermae and Angiospermae. Univ. Arizona brood -rearing potential. Apidologie 11:181 -189. Press. Tucson. 167 pp. Louveaux, J. 1958, 1959. Recherches sur la recolte du pollen par les 100-

5

bd

adee

f9 10- b 9 ee df g

gg ccg9d 9 b t c 9 9 ef 9 c d df f

b d 9 f 15 abb b 9 a d c d d á b d f a d cac é ó a d c b9q df 1.0 - écg bg a f e h f á 9b9 6f bbf ee a 9f9f b 9f dée d c e b b9 f d cccc c d bb g a S d d 9 e d b9 b e f bg a e d d d d d c e ee d a f 9 9 C o e e 9 c a f áá ac f 6 9 b bd 0.1- ac e é 9 a fd 99 9 fbf t 9 b 9 aa de I, a 99 9 c e e 9 e e c c e c ee c d f ff ff e 99 cc 9 c e 9. c dcdc c f c

d d cc 0.01- cf cf f

c ce

ce

f. C C C C

e c c

0.001 dd d 20 30 40 50 60 SPECIES RANK

Figure7.Comparison of the abundance -diversity curves for Honey Bee colony pollen diets. Importance values are the estimated annual percent contribution (P) of each pollen taxon to the colony a and b = colonies 1 and 2, respectively in 1946 at Harpenden, England (Synge, 1947); c = a colony in Wales (Percival, 1947); d = colony A ion 1954 near Manhattan, Kansas (Rashad and Parker, 1956); e, f, and g = average between colonies at Tanque Verde apiary in 1976, 1978, and 1979 respectively 10,000

c

g e a é d g

b g

e 1,000 c 9g b a d d ff a d 9 9 C g ef b d f f ° e ff b C f Q 9 a d d b c f c d ç g 9 dd 9 e c c¿ b b d f9 9 e f g b b dad f f 9 ab e f g a A ab a C ff99 e b a e e ff

e ff e 9 99 9 b 4`a) c ¿bbcc 9 f dtl a D b f

a 8dd dd f 9 9 cb$ é e d d 4 99 99 c p e b 9 aa e e á c bb e t d 99 9g b a A{ c 9 9 a d a e é 8f 9 cc é ff 9 9 9g bb 8 a c á ef b d e ° b a c 9g cc e e e ee 9 g a 9 a f f f ee a gD f aa a ¿ f e d cc cc e

d d C c

d cc d f a

1.0 ffé c e °

e cC C C

e. cc.

d d d f. 0.1 I I I 1 0 20 30 40 510 60 SPECIES RANK Figure 8. Comparison of the abundance- diversity curves for Honey Bee colony pollen diets. Importance values are the estimated annual contribution (g/colony) of each pollen taxon to the colony. Plots are lettered and based on the data in Figure 7. 108 Desert Plants 6 (2) 1984

abeilles (Apis mellifera L.). Ann. Abeille (Paris) 1:113 -118, 197 -221; Manhattan, Kansas area and the influence of weather factors upon 2:13 -111. pollen collection by honey bees. Proc. Tenth Int. Congr. Entomol. Lovell, J. H. 1926. Honey Plants of North America. A. I. Root Co. 4:1037 -1046. Medina. Riedel, S. M. and W. T Wilson. 1967. Pollen collection and behavioral Lowe, C. H. 1964. Arizona landscapes and habitats. In C. H. Lowe (ed.), characteristics of the honey bee at high altitudes. Am. BeeJ. The Vertebrates of Arizona. Univ. Arizona Press. Tucson. 107:10 -12. Markgraf, V. and H. L. D'Antoni. 1978. Pollen Flora of Argentina. Roubik, D. W. 1982a. Obligate necrophagy in a social bee. Science Univ. Arizona Press. Tucson. 208 pp. 21 7:1059 -1060. Maurizio, A. 1941. Über ein massensterben von bienen, verursacht Roubik, D. W. 1982b. Seasonality in colony food storage, brood durch pollen von Ranunculus puberulus Koch. Schweiz. Naturfors. production and adult survivorship: studies of Melipona in tropical Ges. 121:149-150. forest (Hymenoptera: Apidae). f. Kansas Entomol. Soc. 55:789 -800. Maurizio, A. I 945a. Trachtkrankheiten der bienen. I. Vergiftungen bei Roubik, D. W. and C. D. Michener. 1980. The seasonal cycle and einseitiger tracht von rosskastanien. Schweiz. Bienen -Ztg. 1:337 -368. nests of Epicharis zonata, a bee whose cells are below the wet - Maurizio, A. 1945b. Giftige bienenpflanzen. Schweiz. Bienen -Ztg. season water tableHymenoptera, Anthophoridael. Biotropica I: 430 -440. 12:56 -60. Maurizio, A. 1949. Pollenanalytische untersuchungen an honig und Schaffer, W M., D. B. Jensen, D. E. Hobbs, J. Gurevitch, J. R. Todd and pollenhöschen. B. Wird das pollenbild des honigs durch vorgänge M. V. Schaffer. 1979. Competition, foraging energetics, and the cost in der honigblase beeinflusst? Schweiz. Bienen -Ztg. 2:422 -441. of sociality in three species of bees. Ecology 60:976 -987. Maurizio, A. 1953. Weitre untersuchungen an pollenhöschen. Beitrag Schmalzel, R.J.1980. The Diet Breadth of Apis (Hymenoptera: zur erfassung der pollentrachtverhältnisse in verschiedenen ge- Apidae). M.S. Thesis, Univ. Arizona. Tucson. 79 pp. genden der schweiz. Beih. Schweiz. Bienen -Ztg. 2:485 -556. Schremmer, E 1967. Beobachtungen über das sammeln von wind - Maurizio, A. 1960. Bienenbotanik. In A. Budel and E. Herold, ed., pollen durch die honigbiene. Allg. dt. Imkerztg. 1:151 -154. Biene und Bienenzucht. Ehrenwirth Verlag. Munchen. Schwarz, H. E 1948. Stingless bees (Meliponidae) of the Western Maurizio, A. 1964. Mikroskopische und papierchromatographische Hemisphere. Bull. Amer. Mus. Nat. History 90:1 -546. untersuchungen an honig von hummeln, meliponinen und ande- Seely, M. K. 1978. Grassland productivity: the desert end of the curve. ren, zuckerhaltige safte sammelnden insekten. Zeitschrift fur South African Journal of Science 74:295 -297. Bienenforschung 7:98 -110. Severson, D. W. and J. E. Parry. 1981. A chronology of pollen collection May, R. E. 1975. Patterns of species abundance and diversity. pp. by honeybees. Journal of Apicultural Research 20:97 -103. 81 -120. In M. L. Cody andJ. M. Diamond (eds.) Ecology and Shannon, C. E. and W. Weaver. 1949. The Mathematical Theory of Evolution of Communities. Belknap Press of Harvard University Communication. Univ. Illinois Press. Urbana. 117 pp. Press. Cambridge. 545 pp. Shreve, F.1951. Vegetation of the Sonoran Desert. Carnegie Inst. McAndrews, J.H., A. A. Berti and G. Norris. 1973. Key to the Wash. Publ. 591. 192 pp. Quaternary Pollen and Spores of the Great Lakes Region. Life Sci. Simpson, B. 1977. Breeding systems of dominant perennial plants of Misc. Publ. R. Ont. Mus. Toronto. 61 pp. two disjunct warm desert ecosystems. Oecologia 27:203 -226. McGinnies, W G. 1980. Native Desert Plant Flowering Seasons, Smith, E G. 1960. Beekeeping in the Tropics. Tropical Agriculture Tucson Vicinity Office of Arid Lands Studies. Univ. Arizona. Series. Longmans. London. 265 pp. Tucson. Smith, M. V. and A. Adie. 1963. A new design in pollen traps. Can. Bee McClellan, A. R. 1977. Minerals, carbohydrates, and amino acids of J. 74:4, 5, 8. pollens from some woody and herbaceous plants. Ann. Bot. (Lond.) Soehngen, U. and S. C. Jay. 1972. Factors affecting pollen removal from 41:1225 -1232. the honey sac of worker honeybees. J. Apic. Res. 11:3 -8. Michener, C. D. 1979. The biogeography of the bees. Ann. Missouri Solomon, A. M. 1979. Pollen. Pages 41 -54. In R. L. Edmonds, ed., The Bot. Garden 66:277 -347. Ecological Systems Approach. US /IBP Synthesis Series 10. Dowden, Moeller, E E. 1972. Honey bee collection of corn pollen reduced by Hutchinson and Ross, Inc., Stroudsberg. feeding pollen in the hive. Am. Bee J. 112:210 -212. Spencer -Booth, Y. 1960. Feeding pollen, pollen substitutes and pollen Moriya, K. 1960. Studies on honeybee palynology in Sapporo, 1958- supplements to honeybees. Bee World 41:253 -263. 1959. 1. Fac. Sci. Hokkaido Univ. 14:327 -339. Stanley, R. G. and H. E Linskens. 1974. Pollen: Biology, Biochemistry, Muck, O.1936. Bericht der amtlichen untersuchungsstelle für Management. Springer -Verlag, New York. ansteckende bienenkrankheiten ander tierärztlichen hochschule Stapel, A. C. and K. M. Eriksen. 1944. Pollenanalytiske undersgelser Ober das jahr 1935. Wien, Tieräerztl. Monatsschr. 23:168 -173. over honnigbiernes traekplanter. III. Tidsskr. Planteavl 49:303 -318. Muller, C. L. 1948. Die trachtkrankheiten der honigbiene. Tierzucht Synge, A. D. 1947. Pollen collection by honeybees (Apis mellifera). J. 2:117 -120. Anim. Ecol. 16:122 -138. Munz, P. A. 1959. A California Flora. Univ. Calif. Press. Berkeley. Thompson, V. C. 1960. Nectar flow and pollen yield in southwestern Nolan, W. J. 1925. The Brood -rearing Cycle of the Honeybee. U.S.D.A. Arkansas, 1945 -1951. Arkansas Agric. Exp. Sta. Report Series 94. Department Bull. No. 1349. 55 pp. Todd, E. E. and R. K. Bishop. 1940. Trapping honeybee- gathered pollen Noy -Meir, I. 1973. Desert ecosystems: environment and producers. and factors affecting yields. J. Econ. Entomol. 33:866 -870. Ann. Rev. Ecol. Syst. 4:25 -51. Todd, E E. and R. K. Bishop. 1946. The Role of Pollen in the Economy Olsen, L. G. 1975. Pollen Gathering by Honey Bees in Southern of the Hive. U.S. Bur. Ent. and Plant Quar. E -536, Rev., 7 pp. Michigan. M.S. Thesis. Michigan State Univ. East Lansing. Turner, G. J. 1974. Possible transmission of Puccinia polysora by bees. Percival, M. 1947. Pollen collection by Apis mellifera. New Phytolo- Trans. Br. Mycol. Soc. 62:205 -206. gist 46:142 -165. U.S.D.A. 1976. Agricultural Statistics. U. S. Govt. Printing Office. Percival, M.S. 1955. The presentation of pollen in certain angiosperms Wash., D.C. and its collection by Apis mellifera. New Phytologist 54:353 -368. Vansell, G: H. 1926. Buckeye Poisoning of the Honey Bee. Calif. Agrie. Poltev,V.I.1956. (Food toxicosis of bees and their diagnosis] Exp. Sta. Circ. 301. Sixteenth Int. Beekeeping Congr. Vienna. In Russian). Vansell, G. H. and E.E. Todd. 1949. Bee gathered pollen in various Ramírez, W. 1980. Produccion anual de polen por una colmena en el localities on the Pacific coast. Gleanings in Bee Culture 77:18 -21. bosque humedo premontaho costarricense. Agron. Costarr. 4:111 -113. Vogel, S. 1974. Ölblumen und olsammelnde bienen. Trop. Subtrop. Rashad, S. E. and R. L. Parker. 1956. Major pollen sources in the Pf lanzen w. 7:283 -547. O'Neal and Waller Honey Bee Pollen Harvest 109

Wahl, O. 1963. Vergleichende untersuchungen liber nährwert von Whittaker, R. H. 1965. Dominance and diversity in land plant pollen, hefe, soja -mehl, und trockenmilch für die honigbiene (Apis communities. Science 147:250 -260. mellifica); ein beitrag zum eiweiss und vitaminstoffwechsel der Whittaker, R. H. 1972. Evolution and measurement of species biene. Z. Bienenforsch. 6:209 -280. diversity. Taxon 21:213 -251. Waller, G. D. 1980. A modification of the O.A.C. pollen trap. Am. Bee Whittaker, R. H. and W. A. Niering. 1965. Vegetation of the Santa 1. 120:119 -121. Catalina mountains, Arizona: a gradient analysis of the south Waller, G. D., D. M. Caron and G. M. Loper. 1981. Pollen patties slope. Ecology 46:429 -452. maintain brood- rearing when pollen is trapped from honey bee Winston, M. L. 1980. Seasonal patterns of brood rearing and worker colonies. Am. Bee J. 121:101 -103, 105. longevity in the colonies of the Africanized honey bee ¡Hymen- Whitcomb, W 1944. Pollen -trap records and pollination. pp. 161 -162. optera: Apidael in South America. I. Kansas Entomol. Soc. Louisiana Agric. Exp. Sta. Ann. Reports, 1944. 53:157 -165.

Jojoba Simmondsia chinensis (Link.) Schneid. Simmondsiaceae Appendix 1. Key to the pollen groups in Tables 4 and 6. London Rocket Sisymbrium írio L. Cruciferae Manzanita Arctostaphylos spp. Ericaceae Pollen Group Taxa within Pollen Group Family Mesquite Prosopis velutina Woot. Legumínosae Mimosoideae (Mimosa sp. or Calliandra sp.) Leguminosae Acanthus Aniscanthus sp., Beloperone sp. Acanthaceae Mimosa Phoradendron callfornicum Nutt. Loranthaceae Alfalfa Medicago sativa L. Leguminosae Mistletoe Morus spp. esp. M. rubra L. Moraceae Alfilaria Erodium cicutarium (L.) L'Her. Geraniaceae Mulberry Phoenix dactylifera L. and Palmae Aloe Aloe spp. esp. A. saponaria (Ait.) Haw. Liliaceae Palm Washingtonia spp. Amaranth Amaranthus palmeri Wats. Amaranthaceae Paloverde Cercidium spp. and Parkinsonia aculeata L. Leguminosae Ash Fraxinus velutina Torr. Oleaceae Paperflower Psilostrophe cooperi (Gray) Greene Compositae Bindweed Convolvulus spp. Convolvulaceae Lepidium thurberi Wooton Cruciferae Bird -of- Paradise Caesalpinia gilliesü (Wallich ex Hook.) Leguminosae Peppergrass Phacelia spp. esp. P distans Benth. Hydrophyllaceae Benth. Phacelia Bladderpod Lesquerella gordoni (Gray) Wats. Cruciferae (Scorpionweed) Pinus halepensis Mill. Pinaceae Brittlebush Encelia farinosa Gray Compositae Pine Prunus spp. Rosaceae Burro Weed Happlopappus spp. esp. H. tenuisectus Compositae Plum Argemone platyceras Link & Otto. Papaveraceae (Greene) Blake Prickly Poppy Ligustrum sp. Oleaceae Cactus possibly Ferocactus wisliizenii Cactaceae Privet Zygophyllaceae (Engelm.) Britton & Rose Puncture Vine Tribulus terrestres L. spring: Hymenoclea salsola T. & G. var. Compositae California Poppy Eschscholtzia mexicana Greene Papaveraceae Ragweed pentalepis (Rydb.) Benson and Franseria Ceniza Leucophyllum frutescens (Berland( Scrophulariaceae (Ambrosiinae) LM. Johnst. spp. autumn: Ambrosia psilostachya DC., Cheno -am Chenopodiaceae and Amaranthus spp. Chenopodiaceae & Amaranthaceae Hymenoclea monogyra T. &. G. and Xanthium saccharatum Wallr. Cholla Opuntia, subgenus Cylindropuntia Cactaceae Rhus lanced L.f. Anacardiaceae Clammy -weed Polanisia trachysperma Torr. & Gray Capparidaceae Rhus Rosmarinus officinalis L. Labiatae Corn Zea mays L. Gramineae Rosemary Carnegiea gigantea (Engelm.) Britt & Rose. Cactaceae Cottonwood Populus fremontii Wan. Salicaceae Saguaro latropha cardiophylla (Torr.) Muell. Arg. Euphorbiaceae Creosotebush Larrea divaricata Cay. subsp. tridentata Zygophyllaceae Sangre-de-Cristo Gramineae ( Sesse. & Moc. ex DC.) Felg. & Lowe Sorghum Sorghum halepense (L.) Pers. Dasylírion wheeleri Wats. Liliaceae Datura Datura ínoxia Mill. subsp. quinquecuspida Solanaceae Sotol Compositae (Torr.) A. S. Batel. Sow Thistle Sonchus spp. Boerhaavia spp. Nyctaginaceae Desert Broom Baccharis sarothroides Gray Compositae Spiderling Euphorbia spp. or Chamaesyce spp. Euphorbiaceae Desert Marigold Baileya multiradiata Harv. & Gray Compositae Spurge Mentzeha spp., esp. M. pumila (Nutt.) Loasaceae Desert Willow Chilopsis linearis (Cay.) Sweet. Bignoniaceae Stick -leaf Desert Zinnia Zinnia acerosa (DC.) Gray Compositae Torr. & Gray Helianthus spp. and Viguiera sp. Compositae Dock Rumex hymenosepalus Torr. Polygonaceae Sunflower spring: Tamarix pentandra Pall. Tamaricaceae Elderberry Sambucus mexicana Presl. Caprifoliaceae Tamarix (Salt Cedar) autumn: T aphylla (L.) Karst. Elm Ulmus sp. Ulmaceae Heterotheca subaxillaris (Lam.) Compositae Eucalyptus Eucalyptus spp. Myrtaceae Telegraph Plant Britt. & Rusby Evening Primrose Gaura sp. and Oenothera spp. Onagraceae Nicotiana spp. esp. N. trigonophylla Solanaceae Evolvulus Evolvulus sp. Convolvulaceae Tobacco Fiddleneck Amsinckia intermedia Fisch & Meyer Boraginaceae Dunal. Streptanthus arizonicus Wats. Cruciferae Globe Mallow Sphaeralcea spp. Malvaceae Twist Flower Juglans major (Torr.) A. Heller Juglandaceae Grass Gramineae esp. Cyndon dactylon (L.) Pers. Gramineae Walnut Acacia constricta Benth. Leguminosae Hackberry Celtes pallida Torr. and C. reticula Torr. Ulmaceae White -thorn Acacia Compositae Hymenothrix Hymenothrix wislizeni Gray Compositae Wild Lettuce Lactuca spp. Salix spp. Salicaceae Indigo -bush Dalea sp. Leguminosae Willow Lycium sp. Solanaceae Joint -fir Ephedra trifurca Torr. Ephedraceae Wolfberry